SUPREME COURT OF THE UNITED STATES
17 APRIL 2013
(announced the judgment of the Court and delivered the opinion of the Court with respect to Parts I, II–A, II–B, and IV, and an opinion with respect to Parts II–C and III, in which Justice Scalia, Justice Ginsburg, and Justice Kagan join)
In Schmerber v California, 384 U. S. 757 (1966), this Court upheld a warrantless blood test of an individual arrested for driving under the influence of alcohol because the officer “might reasonably have believed that he was confronted with an emergency, in which the delay necessary to obtain a warrant, under the circumstances, threatened the destruction of evidence.” Id., at 770 (internal quotation marks omitted). The question presented here is whether the natural metabolization of alcohol in the bloodstream presents a per se exigency that justifies an exception to the Fourth Amendment’s warrant requirement for nonconsensual blood testing in all drunk-driving cases. We conclude that it does not, and we hold, consistent with general Fourth Amendment principles, that exigency in this context must be determined case by case based on the totality of the circumstances.
While on highway patrol at approximately 2:08 a.m., a Missouri police officer stopped Tyler McNeely’s truck after observing it exceed the posted speed limit and repeatedly cross the centerline. The officer noticed several signs that McNeely was intoxicated, including McNeely’s bloodshot eyes, his slurred speech, and the smell of alcohol on his breath. McNeely acknowledged to the officer that he had consumed “a couple of beers” at a bar, App. 20, and he appeared unsteady on his feet when he exited the truck. After McNeely performed poorly on a battery of field-sobriety tests and declined to use a portable breath-test device to measure his blood alcohol concentration (BAC), the officer placed him under arrest.
The officer began to transport McNeely to the station house. But when McNeely indicated that he would again refuse to provide a breath sample, the officer changed course and took McNeely to a nearby hospital for blood testing. The officer did not attempt to secure a warrant. Upon arrival at the hospital, the officer asked McNeely whether he would consent to a blood test. Reading from a standard implied consent form, the officer explained to McNeely that under state law refusal to submit voluntarily to the test would lead to the immediate revocation of his driver’s license for one year and could be used against him in a future prosecution. See Mo. Ann. Stat. §§577.020.1, 577.041 (West 2011). McNeely nonetheless refused. The officer then directed a hospital lab technician to take a blood sample, and the sample was secured at approximately 2:35 a.m. Subsequent laboratory testing measured McNeely’s BAC at 0.154 percent, which was well above the legal limit of 0.08 percent. See §577.012.1.
McNeely was charged with driving while intoxicated (DWI), in violation of §577.010. He moved to suppress the results of the blood test, arguing in relevant part that, under the circumstances, taking his blood for chemical testing without first obtaining a search warrant violated his rights under the Fourth Amendment. The trial court agreed. It concluded that the exigency exception to the warrant requirement did not apply because, apart from the fact that “[a]s in all cases involving intoxication, [McNeely’s] blood alcohol was being metabolized by his liver,” there were no circumstances suggesting the officer faced an emergency in which he could not practicably obtain a warrant. No. 10CG–CR01849–01 (Cir. Ct. Cape Giradeau Cty., Mo., Div. II, Mar. 3, 2011), App. to Pet. for Cert. 43a. On appeal, the Missouri Court of Appeals stated an intention to reverse but transferred the case directly to the Missouri Supreme Court. No. ED 96402 (June 21, 2011), id., at 24a.
The Missouri Supreme Court affirmed. 358 S. W. 3d 65 (2012) (per curiam). Recognizing that this Court’s decision in Schmerber v California, 384 U. S. 757, “provide[d] the backdrop” to its analysis, the Missouri Supreme Court held that “Schmerber directs lower courts to engage in a totality of the circumstances analysis when determining whether exigency permits a nonconsensual, warrantless blood draw.” 358 S. W. 3d, at 69, 74. The court further concluded that Schmerber “requires more than the mere dissipation of blood-alcohol evidence to support a warrantless blood draw in an alcohol-related case.” 358 S. W. 3d, at 70. According to the court, exigency depends heavily on the existence of additional “‘special facts,’” such as whether an officer was delayed by the need to investigate an accident and transport an injured suspect to the hospital, as had been the case in Schmerber. 358 S. W. 3d, at 70, 74. Finding that this was “unquestionably a routine DWI case” in which no factors other than the natural dissipation of blood-alcohol suggested that there was an emergency, the court held that the nonconsensual warrantless blood draw violated McNeely’s Fourth Amendment right to be free from unreasonable searches of his person. Id., at 74–75.
We granted certiorari to resolve a split of authority on the question whether the natural dissipation of alcohol in the bloodstream establishes a per se exigency that suffices on its own to justify an exception to the warrant requirement for nonconsensual blood testing in drunk-driving investigations. See 567 U. S. ___ (2012). We now affirm.
The Fourth Amendment provides in relevant part that “[t]he right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated, and no Warrants shall issue, but upon probable cause.” Our cases have held that a warrantless search of the person is reasonable only if it falls within a recognized exception. See, e.g., United States v Robinson, 414 U. S. 218, 224 (1973). That principle applies to the type of search at issue in this case, which involved a compelled physical intrusion beneath McNeely’s skin and into his veins to obtain a sample of his blood for use as evidence in a criminal investigation. Such an invasion of bodily integrity implicates an individual’s “most personal and deep-rooted expectations of privacy.” Winston v Lee, 470 U. S. 753, 760 (1985); see also Skinner v Railway Labor Executives’ Assn., 489 U. S. 602, 616 (1989).
We first considered the Fourth Amendment restrictions on such searches in Schmerber, where, as in this case, a blood sample was drawn from a defendant suspected of driving while under the influence of alcohol. 384 U. S., at 758. Noting that “[s]earch warrants are ordinarily required for searches of dwellings,” we reasoned that “absent an emergency, no less could be required where intrusions into the human body are concerned,” even when the search was conducted following a lawful arrest. Id., at 770. We explained that the importance of requiring authorization by a “‘neutral and detached magistrate’” before allowing a law enforcement officer to “invade another’s body in search of evidence of guilt is indisputable and great.” Ibid. (quoting Johnson v United States, 333 U. S. 10–14 (1948)).
As noted, the warrant requirement is subject to exceptions. “One well-recognized exception,” and the one at issue in this case, “applies when the exigencies of the situation make the needs of law enforcement so compelling that a warrantless search is objectively reasonable under the Fourth Amendment.” Kentucky v King, 563 U. S. ___, ___ (2011) (slip op., at 6) (internal quotation marks and brackets omitted). A variety of circumstances may give rise to an exigency sufficient to justify a warrantless search, including law enforcement’s need to provide emergency assistance to an occupant of a home, Michigan v Fisher, 558 U. S. 45–48 (2009) (per curiam), engage in “hot pursuit” of a fleeing suspect, United States v Santana, 427 U. S. 38–43 (1976), or enter a burning building to put out a fire and investigate its cause, Michigan v Tyler, 436 U. S. 499–510 (1978). As is relevant here, we have also recognized that in some circumstances law enforcement officers may conduct a search without a warrant to prevent the imminent destruction of evidence. See Cupp v Murphy, 412 U. S. 291, 296 (1973); Ker v California, 374 U. S. 23–41 (1963) (plurality opinion). While these contexts do not necessarily involve equivalent dangers, in each a warrantless search is potentially reasonable because “there is compelling need for official action and no time to secure a warrant.” Tyler, 436 U. S., at 509.
To determine whether a law enforcement officer faced an emergency that justified acting without a warrant, this Court looks to the totality of circumstances. See Brigham City v Stuart, 547 U. S. 398, 406 (2006) (finding officers’ entry into a home to provide emergency assistance “plainly reasonable under the circumstances”); Illinois v McArthur, 531 U. S. 326, 331 (2001) (concluding that a warrantless seizure of a person to prevent him from returning to his trailer to destroy hidden contraband was reasonable “[i]n the circumstances of the case before us” due to exigency); Cupp, 412 U. S., at 296 (holding that a limited warrantless search of a suspect’s fingernails to preserve evidence that the suspect was trying to rub off was justified “[o]n the facts of this case”); see also Richards v Wisconsin, 520 U. S. 385–396 (1997) (rejecting a per se exception to the knock-and-announce requirement for felony drug investigations based on presumed exigency, and requiring instead evaluation of police conduct “in a particular case”). We apply this “finely tuned approach” to Fourth Amendment reasonableness in this context because the police action at issue lacks “the traditional justification that .... a warrant .... provides.” Atwater v Lago Vista, 532 U. S. 318, 347, n. 16 (2001). Absent that established justification, “the fact-specific nature of the reasonableness inquiry,” Ohio v Robinette, 519 U. S. 33, 39 (1996) , demands that we evaluate each case of alleged exigency based “on its own facts and circumstances.” Go-Bart Importing Co. v United States, 282 U. S. 344, 357 (1931).
Our decision in Schmerber applied this totality of the circumstances approach. In that case, the petitioner had suffered injuries in an automobile accident and was taken to the hospital. 384 U. S., at 758. While he was there receiving treatment, a police officer arrested the petitioner for driving while under the influence of alcohol and ordered a blood test over his objection. Id., at 758–759. After explaining that the warrant requirement applied generally to searches that intrude into the human body, we concluded that the warrantless blood test “in the present case” was nonetheless permissible because the officer “might reasonably have believed that he was confronted with an emergency, in which the delay necessary to obtain a warrant, under the circumstances, threatened ‘the destruction of evidence.’” Id., at 770 (quoting Preston v United States, 376 U. S. 364, 367 (1964)).
In support of that conclusion, we observed that evidence could have been lost because “the percentage of alcohol in the blood begins to diminish shortly after drinking stops, as the body functions to eliminate it from the system.” 384 U. S., at 770. We added that “[p]articularly in a case such as this, where time had to be taken to bring the accused to a hospital and to investigate the scene of the accident, there was no time to seek out a magistrate and secure a warrant.” Id., at 770–771. “Given these special facts,” we found that it was appropriate for the police to act without a warrant. Id., at 771. We further held that the blood test at issue was a reasonable way to recover the evidence because it was highly effective, “involve[d] virtually no risk, trauma, or pain,” and was conducted in a reasonable fashion “by a physician in a hospital environment according to accepted medical practices.” Ibid. And in conclusion, we noted that our judgment that there had been no Fourth Amendment violation was strictly based “on the facts of the present record.” Id., at 772.
Thus, our analysis in Schmerber fits comfortably within our case law applying the exigent circumstances exception. In finding the warrantless blood test reasonable in Schmerber, we considered all of the facts and circumstances of the particular case and carefully based our holding on those specific facts.
The State properly recognizes that the reasonableness of a warrantless search under the exigency exception to the warrant requirement must be evaluated based on the totality of the circumstances. Brief for Petitioner 28–29. But the State nevertheless seeks a per se rule for blood testing in drunk-driving cases. The State contends that whenever an officer has probable cause to believe an individual has been driving under the influence of alcohol, exigent circumstances will necessarily exist because BAC evidence is inherently evanescent. As a result, the State claims that so long as the officer has probable cause and the blood test is conducted in a reasonable manner, it is categorically reasonable for law enforcement to obtain the blood sample without a warrant.
It is true that as a result of the human body’s natural metabolic processes, the alcohol level in a person’s blood begins to dissipate once the alcohol is fully absorbed and continues to decline until the alcohol is eliminated. See Skinner, 489 U. S., at 623; Schmerber, 384 U. S., at 770–771. Testimony before the trial court in this case indicated that the percentage of alcohol in an individual’s blood typically decreases by approximately 0.015 percent to 0.02 percent per hour once the alcohol has been fully absorbed. App. 47. More precise calculations of the rate at which alcohol dissipates depend on various individual characteristics (such as weight, gender, and alcohol tolerance) and the circumstances in which the alcohol was consumed. See Stripp, Forensic and Clinical Issues in Alcohol Analysis, in Forensic Chemistry Handbook 437–441 (L. Kobilinsky ed. 2012). Regardless of the exact elimination rate, it is sufficient for our purposes to note that because an individual’s alcohol level gradually declines soon after he stops drinking, a significant delay in testing will negatively affect the probative value of the results. This fact was essential to our holding in Schmerber, as we recognized that, under the circumstances, further delay in order to secure a warrant after the time spent investigating the scene of the accident and transporting the injured suspect to the hospital to receive treatment would have threatened the destruction of evidence. 384 U. S., at 770–771.
But it does not follow that we should depart from careful case-by-case assessment of exigency and adopt the categorical rule proposed by the State and its amici. In those drunk-driving investigations where police officers can reasonably obtain a warrant before a blood sample can be drawn without significantly undermining the efficacy of the search, the Fourth Amendment mandates that they do so. See McDonald v United States, 335 U. S. 451, 456 (1948) (“We cannot .... excuse the absence of a search warrant without a showing by those who seek exemption from the constitutional mandate that the exigencies of the situation made [the search] imperative”). We do not doubt that some circumstances will make obtaining a warrant impractical such that the dissipation of alcohol from the bloodstream will support an exigency justifying a properly conducted warrantless blood test. That, however, is a reason to decide each case on its facts, as we did in Schmerber, not to accept the “considerable overgeneralization” that a per se rule would reflect. Richards, 520 U. S., at 393.
The context of blood testing is different in critical respects from other destruction-of-evidence cases in which the police are truly confronted with a “‘now or never’” situation. Roaden v Kentucky, 413 U. S. 496, 505 (1973). In contrast to, for example, circumstances in which the suspect has control over easily disposable evidence, see Georgia v Randolph, 547 U. S. 103, n. 6 (2006); Cupp, 412 U. S., at 296, BAC evidence from a drunk-driving suspect naturally dissipates over time in a gradual and relatively predictable manner. Moreover, because a police officer must typically transport a drunk-driving suspect to a medical facility and obtain the assistance of someone with appropriate medical training before conducting a blood test, some delay between the time of the arrest or accident and the time of the test is inevitable regardless of whether police officers are required to obtain a warrant. See State v Shriner, 751 N. W. 2d 538, 554 (Minn. 2008) (Meyer, J., dissenting). This reality undermines the force of the State’s contention, endorsed by the dissent, see post, at 3 (opinion of Thomas, J.), that we should recognize a categorical exception to the warrant requirement because BAC evidence “is actively being destroyed with every minute that passes.” Brief for Petitioner 27. Consider, for example, a situation in which the warrant process will not significantly increase the delay before the blood test is conducted because an officer can take steps to secure a warrant while the suspect is being transported to a medical facility by another officer. In such a circumstance, there would be no plausible justification for an exception to the warrant requirement.
The State’s proposed per se rule also fails to account for advances in the 47 years since Schmerber was decided that allow for the more expeditious processing of warrant applications, particularly in contexts like drunk-driving investigations where the evidence offered to establish probable cause is simple. The Federal Rules of Criminal Procedure were amended in 1977 to permit federal magistrate judges to issue a warrant based on sworn testimony communicated by telephone. See 91Stat. 319. As amended, the law now allows a federal magistrate judge to consider “information communicated by telephone or other reliable electronic means.” Fed. Rule Crim. Proc. 4.1. States have also innovated. Well over a majority of States allow police officers or prosecutors to apply for search warrants remotely through various means, including telephonic or radio communication, electronic communication such as e-mail, and video conferencing. And in addition to technology-based developments, jurisdictions have found other ways to streamline the warrant process, such as by using standard-form warrant applications for drunk-driving investigations.
We by no means claim that telecommunications innovations have, will, or should eliminate all delay from the warrant-application process. Warrants inevitably take some time for police officers or prosecutors to complete and for magistrate judges to review. Telephonic and electronic warrants may still require officers to follow time-consuming formalities designed to create an adequate record, such as preparing a duplicate warrant before calling the magistrate judge. See Fed. Rule Crim. Proc. 4.1(b)(3). And improvements in communications technology do not guarantee that a magistrate judge will be available when an officer needs a warrant after making a late-night arrest. But technological developments that enable police officers to secure warrants more quickly, and do so without undermining the neutral magistrate judge’s essential role as a check on police discretion, are relevant to an assessment of exigency. That is particularly so in this context, where BAC evidence is lost gradually and relatively predictably.
Of course, there are important countervailing concerns. While experts can work backwards from the BAC at the time the sample was taken to determine the BAC at the time of the alleged offence, longer intervals may raise questions about the accuracy of the calculation. For that reason, exigent circumstances justifying a warrantless blood sample may arise in the regular course of law enforcement due to delays from the warrant application process. But adopting the State’s per se approach would improperly ignore the current and future technological developments in warrant procedures, and might well diminish the incentive for jurisdictions “to pursue progressive approaches to warrant acquisition that preserve the protections afforded by the warrant while meeting the legitimate interests of law enforcement.” State v Rodriguez, 2007 UT 15, ¶46, 156 P. 3d 771, 779.
In short, while the natural dissipation of alcohol in the blood may support a finding of exigency in a specific case, as it did in Schmerber, it does not do so categorically. Whether a warrantless blood test of a drunk-driving suspect is reasonable must be determined case by case based on the totality of the circumstances.
In an opinion concurring in part and dissenting in part, The Chief Justice agrees that the State’s proposed per se rule is overbroad because “[f]or exigent circumstances to justify a warrantless search .... there must .... be ‘no time to secure a warrant.’” Post, at 6 (quoting Tyler, 436 U. S., at 509). But The Chief Justice then goes on to suggest his own categorical rule under which a warrantless blood draw is permissible if the officer could not secure a warrant (or reasonably believed he could not secure a warrant) in the time it takes to transport the suspect to a hospital or similar facility and obtain medical assistance. Post, at 8–9. Although we agree that delay inherent to the blood-testing process is relevant to evaluating exigency, see supra, at 10, we decline to substitute The Chief Justice’s modified per se rule for our traditional totality of the circumstances analysis.
For one thing, making exigency completely dependent on the window of time between an arrest and a blood test produces odd consequences. Under The Chief Justice’s rule, if a police officer serendipitously stops a suspect near an emergency room, the officer may conduct a nonconsensual warrantless blood draw even if all agree that a warrant could be obtained with very little delay under the circumstances (perhaps with far less delay than an average ride to the hospital in the jurisdiction). The rule would also distort law enforcement incentives. As with the State’s per se rule, The Chief Justice’s rule might discourage efforts to expedite the warrant process because it categorically authorizes warrantless blood draws so long as it takes more time to secure a warrant than to obtain medical assistance. On the flip side, making the requirement of independent judicial oversight turn exclusively on the amount of time that elapses between an arrest and BAC testing could induce police departments and individual officers to minimize testing delay to the detriment of other values. The Chief Justice correctly observes that “[t]his case involves medical personnel drawing blood at a medical facility, not police officers doing so by the side of the road.” Post, at 6–7, n. 2. But The Chief Justice does not say that roadside blood draws are necessarily unreasonable, and if we accepted The Chief Justice’s approach, they would become a more attractive option for the police.
The remaining arguments advanced in support of a per se exigency rule are unpersuasive.
The State and several of its amici, including the United States, express concern that a case-by-case approach to exigency will not provide adequate guidance to law enforcement officers deciding whether to conduct a blood test of a drunk-driving suspect without a warrant. The Chief Justice and the dissent also raise this concern. See post, at 1, 9–10 (opinion of Roberts, C. J.); post, at 5–7 (opinion of Thomas, J.). While the desire for a bright-line rule is understandable, the Fourth Amendment will not tolerate adoption of an overly broad categorical approach that would dilute the warrant requirement in a context where significant privacy interests are at stake. Moreover, a case-by-case approach is hardly unique within our Fourth Amendment jurisprudence. Numerous police actions are judged based on fact-intensive, totality of the circumstances analyses rather than according to categorical rules, including in situations that are more likely to require police officers to make difficult split-second judgments. See, e.g., Illinois v Wardlow, 528 U. S. 119–125 (2000) (whether an officer has reasonable suspicion to make an investigative stop and to pat down a suspect for weapons under Terry v Ohio, 392 U. S. 1 (1968)); Robinette, 519 U. S., at 39–40 (whether valid consent has been given to search); Tennessee v Garner, 471 U. S. 1–9, 20 (1985) (whether force used to effectuate a seizure, including deadly force, is reasonable). As in those contexts, we see no valid substitute for careful case-by-case evaluation of reasonableness here.
Next, the State and the United States contend that the privacy interest implicated by blood draws of drunk-driving suspects is relatively minimal. That is so, they claim, both because motorists have a diminished expectation of privacy and because our cases have repeatedly indicated that blood testing is commonplace in society and typically involves “virtually no risk, trauma, or pain.” Schmerber, 384 U. S., at 771. See also post, at 3, and n. 1 (opinion of Thomas, J.).
But the fact that people are “accorded less privacy in .... automobiles because of th[e] compelling governmental need for regulation,” California v Carney, 471 U. S. 386, 392 (1985) , does not diminish a motorist’s privacy interest in preventing an agent of the government from piercing his skin. As to the nature of a blood test conducted in a medical setting by trained personnel, it is concededly less intrusive than other bodily invasions we have found unreasonable. See Winston, 470 U. S., at 759–766 (surgery to remove a bullet); Rochin v California, 342 U. S. 165–174 (1952) (induced vomiting to extract narcotics capsules ingested by a suspect violated the Due Process Clause). For that reason, we have held that medically drawn blood tests are reasonable in appropriate circumstances. See Skinner, 489 U. S., at 618–633 (upholding warrantless blood testing of railroad employees involved in certain train accidents under the “special needs” doctrine); Schmerber, 384 U. S., at 770–772. We have never retreated, however, from our recognition that any compelled intrusion into the human body implicates significant, constitutionally protected privacy interests.
Finally, the State and its amici point to the compelling governmental interest in combating drunk driving and contend that prompt BAC testing, including through blood testing, is vital to pursuit of that interest. They argue that is particularly so because, in addition to laws that make it illegal to operate a motor vehicle under the influence of alcohol, all 50 States and the District of Columbia have enacted laws that make it per se unlawful to operate a motor vehicle with a BAC of over 0.08 percent. See National Highway Traffic Safety Admin. (NHTSA), Alcohol and Highway Safety: A Review of the State of Knowledge 167 (No. 811374, Mar. 2011) (NHTSA Review).  To enforce these provisions, they reasonably assert, accurate BAC evidence is critical. See also post, at 4–5 (opinion of Roberts, C. J.); post, at 4–5 (opinion of Thomas, J.).
“No one can seriously dispute the magnitude of the drunken driving problem or the States’ interest in eradicating it.” Michigan Dept. of State Police v Sitz, 496 U. S. 444, 451 (1990). Certainly we do not. While some progress has been made, drunk driving continues to exact a terrible toll on our society. See NHTSA, Traffic Safety Facts, 2011 Data 1 (No. 811700, Dec. 2012) (reporting that 9,878 people were killed in alcohol-impaired driving crashes in 2011, an average of one fatality every 53 minutes).
But the general importance of the government’s interest in this area does not justify departing from the warrant requirement without showing exigent circumstances that make securing a warrant impractical in a particular case. To the extent that the State and its amici contend that applying the traditional Fourth Amendment totality-of-the-circumstances analysis to determine whether an exigency justified a warrantless search will undermine the governmental interest in preventing and prosecuting drunk-driving offences, we are not convinced.
As an initial matter, States have a broad range of legal tools to enforce their drunk-driving laws and to secure BAC evidence without undertaking warrantless nonconsensual blood draws. For example, all 50 States have adopted implied consent laws that require motorists, as a condition of operating a motor vehicle within the State, to consent to BAC testing if they are arrested or otherwise detained on suspicion of a drunk-driving offence. See NHTSA Review 173; supra, at 2 (describing Missouri’s implied consent law). Such laws impose significant consequences when a motorist withdraws consent; typically the motorist’s driver’s license is immediately suspended or revoked, and most States allow the motorist’s refusal to take a BAC test to be used as evidence against him in a subsequent criminal prosecution. See NHTSA Review 173–175; see also South Dakota v Neville, 459 U. S. 553–564 (1983) (holding that the use of such an adverse inference does not violate the Fifth Amendment right against self-incrimination).
It is also notable that a majority of States either place significant restrictions on when police officers may obtain a blood sample despite a suspect’s refusal (often limiting testing to cases involving an accident resulting in death or serious bodily injury) or prohibit nonconsensual blood tests altogether.  Among these States, several lift restrictions on nonconsensual blood testing if law enforcement officers first obtain a search warrant or similar court order. Cf. Bullcoming v New Mexico, 564 U. S. ___, ___ (2011) (slip op., at 3) (noting that the blood test was obtained pursuant to a warrant after the petitioner refused a breath test). We are aware of no evidence indicating that restrictions on nonconsensual blood testing have compromised drunk-driving enforcement efforts in the States that have them. And in fact, field studies in States that permit nonconsensual blood testing pursuant to a warrant have suggested that, although warrants do impose administrative burdens, their use can reduce breath-test-refusal rates and improve law enforcement’s ability to recover BAC evidence. See NHTSA, Use of Warrants for Breath Test Refusal: Case Studies 36–38 (No. 810852, Oct. 2007).
To be sure, “States [may] choos[e] to protect privacy beyond the level that the Fourth Amendment requires.” Virginia v Moore, 553 U. S. 164, 171 (2008). But wide-spread state restrictions on nonconsensual blood testing provide further support for our recognition that compelled blood draws implicate a significant privacy interest. They also strongly suggest that our ruling today will not “severely hamper effective law enforcement.” Garner, 471 U. S., at 19.
The State argued before this Court that the fact that alcohol is naturally metabolized by the human body creates an exigent circumstance in every case. The State did not argue that there were exigent circumstances in this particular case because a warrant could not have been obtained within a reasonable amount of time. In his testimony before the trial court, the arresting officer did not identify any other factors that would suggest he faced an emergency or unusual delay in securing a warrant. App. 40. He testified that he made no effort to obtain a search warrant before conducting the blood draw even though he was “sure” a prosecuting attorney was on call and even though he had no reason to believe that a magistrate judge would have been unavailable. Id., at 39, 41–42. The officer also acknowledged that he had obtained search warrants before taking blood samples in the past without difficulty. Id., at 42. He explained that he elected to forgo a warrant application in this case only because he believed it was not legally necessary to obtain a warrant. Id., at 39–40. Based on this testimony, the trial court concluded that there was no exigency and specifically found that, although the arrest took place in the middle of the night, “a prosecutor was readily available to apply for a search warrant and a judge was readily available to issue a warrant.” App. to Pet. for Cert. 43a.
The Missouri Supreme Court in turn affirmed that judgment, holding first that the dissipation of alcohol did not establish a per se exigency, and second that the State could not otherwise satisfy its burden of establishing exigent circumstances. 358 S. W. 3d, at 70, 74–75. In petitioning for certiorari to this Court, the State challenged only the first holding; it did not separately contend that the warrantless blood test was reasonable regardless of whether the natural dissipation of alcohol in a suspect’s blood categorically justifies dispensing with the warrant requirement. See Pet. for Cert. i.
Here and in its own courts the State based its case on an insistence that a driver who declines to submit to testing after being arrested for driving under the influence of alcohol is always subject to a nonconsensual blood test without any precondition for a warrant. That is incorrect.
Although the Missouri Supreme Court referred to this case as “unquestionably a routine DWI case,” 358 S. W. 3d, at 74, the fact that a particular drunk-driving stop is “routine” in the sense that it does not involve “‘special facts,’” ibid., such as the need for the police to attend to a car accident, does not mean a warrant is required. Other factors present in an ordinary traffic stop, such as the procedures in place for obtaining a warrant or the availability of a magistrate judge, may affect whether the police can obtain a warrant in an expeditious way and therefore may establish an exigency that permits a warrantless search. The relevant factors in determining whether a warrantless search is reasonable, including the practical problems of obtaining a warrant within a timeframe that still preserves the opportunity to obtain reliable evidence, will no doubt vary depending upon the circumstances in the case.
Because this case was argued on the broad proposition that drunk-driving cases present a per se exigency, the arguments and the record do not provide the Court with an adequate analytic framework for a detailed discussion of all the relevant factors that can be taken into account in determining the reasonableness of acting without a warrant. It suffices to say that the metabolization of alcohol in the bloodstream and the ensuing loss of evidence are among the factors that must be considered in deciding whether a warrant is required. No doubt, given the large number of arrests for this offence in different jurisdictions nationwide, cases will arise when anticipated delays in obtaining a warrant will justify a blood test without judicial authorization, for in every case the law must be concerned that evidence is being destroyed. But that inquiry ought not to be pursued here where the question is not properly before this Court. Having rejected the sole argument presented to us challenging the Missouri Supreme Court’s decision, we affirm its judgment.
* * *
We hold that in drunk-driving investigations, the natural dissipation of alcohol in the bloodstream does not constitute an exigency in every case sufficient to justify conducting a blood test without a warrant.
The judgment of the Missouri Supreme Court is affirmed.
It is so ordered.
(concurring in part)
I join Parts I, II–A, II–B, and IV of the opinion for the Court.
For the reasons stated below this case does not call for the Court to consider in detail the issue discussed in Part II–C and the separate opinion by The Chief Justice.
As to Part III, much that is noted with respect to the statistical and survey data will be of relevance when this issue is explored in later cases. The repeated insistence in Part III that every case be determined by its own circumstances is correct, of course, as a general proposition; yet it ought not to be interpreted to indicate this question is not susceptible of rules and guidelines that can give important, practical instruction to arresting officers, instruction that in any number of instances would allow a warrantless blood test in order to preserve the critical evidence.
States and other governmental entities which enforce the driving laws can adopt rules, procedures, and protocols that meet the reasonableness requirements of the Fourth Amendment and give helpful guidance to law enforcement officials. And this Court, in due course, may find it appropriate and necessary to consider a case permitting it to provide more guidance than it undertakes to give today.
As the opinion of the Court is correct to note, the instant case, by reason of the way in which it was presented and decided in the state courts, does not provide a framework where it is prudent to hold any more than that always dispensing with a warrant for a blood test when a driver is arrested for being under the influence of alcohol is inconsistent with the Fourth Amendment.
Chief Justice Roberts
(with whom Justice Breyer and Justice Alito join, concurring in part and dissenting in part)
A police officer reading this Court’s opinion would have no idea – no idea – what the Fourth Amendment requires of him, once he decides to obtain a blood sample from a drunk driving suspect who has refused a breathalyzer test. I have no quarrel with the Court’s “totality of the circumstances” approach as a general matter; that is what our cases require. But the circumstances in drunk driving cases are often typical, and the Court should be able to offer guidance on how police should handle cases like the one before us.
In my view, the proper rule is straightforward. Our cases establish that there is an exigent circumstances exception to the warrant requirement. That exception applies when there is a compelling need to prevent the imminent destruction of important evidence, and there is no time to obtain a warrant. The natural dissipation of alcohol in the bloodstream constitutes not only the imminent but ongoing destruction of critical evidence. That would qualify as an exigent circumstance, except that there may be time to secure a warrant before blood can be drawn. If there is, an officer must seek a warrant. If an officer could reasonably conclude that there is not, the exigent circumstances exception applies by its terms, and the blood may be drawn without a warrant.
The Fourth Amendment provides:
The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated, and no Warrants shall issue, but upon probable cause, supported by Oath or affirmation, and particularly describing the place to be searched, and the persons or things to be seized.
That language does not state that warrants are required prior to searches, but this Court has long held that warrants must generally be obtained. See Kentucky v King, 563 U. S. ___, ___ (2011) (slip op., at 5). We have also held that bodily intrusions like blood draws constitute searches and are subject to the warrant requirement. See Schmerber v California, 384 U. S. 757, 767, 770 (1966).
However, “the ultimate touchstone of the Fourth Amendment is ‘reasonableness,’” Brigham City v Stuart, 547 U. S. 398, 403 (2006), and thus “the warrant requirement is subject to certain reasonable exceptions,” King, 563 U. S., at ___ (slip op., at 6). One of those exceptions is known as the “exigent circumstances exception,” which “applies when the exigencies of the situation make the needs of law enforcement so compelling that a warrantless search is objectively reasonable under the Fourth Amendment.” Ibid. (internal quotation marks and alterations omitted).
Within the exigent circumstances exception, we have identified several sets of exigent circumstances excusing the need for a warrant. For example, there is an emergency aid exception to the warrant requirement. In Brigham City, supra, at 403, we held that “law enforcement officers may enter a home without a warrant to render emergency assistance to an injured occupant or to protect an occupant from imminent injury.” There is also a fire exception to the warrant requirement. In Michigan v Tyler, 436 U. S. 499, 509 (1978) , we held that “[a] burning building clearly presents an exigency of sufficient proportions to render a warrantless entry ‘reasonable.’” And there is a hot pur-suit exception to the warrant requirement as well. In United States v Santana, 427 U. S. 38 (1976), and Warden, Md. Penitentiary v Hayden, 387 U. S. 294 (1967) , we recognized “the right of police, who had probable cause to believe that an armed robber had entered a house a few minutes before, to make a warrantless entry to arrest the robber and to search for weapons.” Santana, supra, at 42. In each of these cases, the requirement that we base our decision on the “totality of the circumstances” has not prevented us from spelling out a general rule for the police to follow.
The exigency exception most on point here is the one for imminent destruction of evidence. We have affirmed on several occasions that “law enforcement officers may make a warrantless entry onto private property .... to prevent the imminent destruction of evidence.” Brigham City, supra, at 403 (citing Ker v California, 374 U. S. 23, 40 (1963) (plurality opinion)); see also, e.g., King, supra, at ___ (slip op., at 6). For example, in Ker, the police had reason to believe that the defendant was in possession of marijuana and was expecting police pursuit. We upheld the officers’ warrantless entry into the defendant’s home, with the plurality explaining that the drugs “could be quickly and easily destroyed” or “distributed or hidden before a warrant could be obtained at that time of night.” 374 U. S., at 40, 42.
As an overarching principle, we have held that if there is a “compelling need for official action and no time to secure a warrant,” the warrant requirement may be excused. Tyler, supra, at 509. The question here is whether and how this principle applies in the typical case of a police officer stopping a driver on suspicion of drunk driving.
The reasonable belief that critical evidence is being destroyed gives rise to a compelling need for blood draws in cases like this one. Here, in fact, there is not simply a belief that any alcohol in the bloodstream will be destroyed; it is a biological certainty. Alcohol dissipates from the bloodstream at a rate of 0.01 percent to 0.025 percent per hour. Stripp, Forensic and Clinical Issues in Alcohol Analysis, in Forensic Chemistry Handbook 440 (L. Kobilinsky ed. 2012). Evidence is literally disappearing by the minute. That certainty makes this case an even stronger one than usual for application of the exigent circumstances exception.
And that evidence is important. A serious and deadly crime is at issue. According to the Department of Transportation, in 2011, one person died every 53 minutes due to drinking and driving. National Highway Traffic Safety Admin. (NHTSA), Traffic Safety Facts, 2011 Data 1 (No. 811700, Dec. 2012). No surprise then that drinking and driving is punished severely, including with jail time. See generally Dept. of Justice, Bureau of Justice Statistics, L. Maruschak, Special Report, DWI Offenders under Correctional Supervision (1999). McNeely, for instance, faces up to four years in prison. See App. 22–23 (citing Mo. Ann. Stat. §§558.011, 577.010, 577.023 (West 2011)).
Evidence of a driver’s blood alcohol concentration (BAC) is crucial to obtain convictions for such crimes. All 50 States and the District of Columbia have laws providing that it is per se illegal to drive with a BAC of 0.08 percent or higher. Most States also have laws establishing additional penalties for drivers who drive with a “high BAC,” often defined as 0.15 percent or above. NHTSA, Digest of Impaired Driving and Selected Beverage Control Laws, pp. vii, x–xviii (No. 811673, Oct. 2012). BAC evidence clearly matters. And when drivers refuse breathalyzers, as McNeely did here, a blood draw becomes necessary to obtain that evidence.
The need to prevent the imminent destruction of BAC evidence is no less compelling because the incriminating alcohol dissipates over a limited period of time, rather than all at once. As noted, the concentration of alcohol can make a difference not only between guilt and innocence, but between different crimes and different degrees of punishment. The officer is unlikely to know precisely when the suspect consumed alcohol or how much; all he knows is that critical evidence is being steadily lost. Fire can spread gradually, but that does not lessen the need and right of the officers to respond immediately. See Tyler, supra.
McNeely contends that there is no compelling need for a warrantless blood draw, because if there is some alcohol left in the blood by the time a warrant is obtained, the State can use math and science to work backwards and identify a defendant’s BAC at the time he was driving. See Brief for Respondent 44–46. But that’s not good enough. We have indicated that exigent circumstances justify warrantless entry when drugs are about to be flushed down the toilet. See, e.g., King, 563 U. S., at ___–___ (slip op., at 7–8). We have not said that, because there could well be drug paraphernalia elsewhere in the home, or because a defendant’s co-conspirator might testify to the amount of drugs involved, the drugs themselves are not crucial and there is no compelling need for warrantless entry.
The same approach should govern here. There is a compelling need to search because alcohol – the nearly conclusive evidence of a serious crime – is dissipating from the bloodstream. The need is no less compelling because the police might be able to acquire second-best evidence some other way.
For exigent circumstances to justify a warrantless search, however, there must also be “no time to secure a warrant.” Tyler, 436 U. S., at 509; see Schmerber, 384 U. S., at 771 (warrantless search legal when “there was no time to seek out a magistrate and secure a warrant”). In this respect, obtaining a blood sample from a suspected drunk driver differs from other exigent circumstances cases.
Importantly, there is typically delay between the moment a drunk driver is stopped and the time his blood can be drawn. Drunk drivers often end up in an emergency room, but they are not usually pulled over in front of one. In most exigent circumstances situations, police are just outside the door to a home. Inside, evidence is about to be destroyed, a person is about to be injured, or a fire has broken out. Police can enter promptly and must do so to respond effectively to the emergency. But when police pull a person over on suspicion of drinking and driving, they cannot test his blood right away. There is a time-consuming obstacle to their search, in the form of a trip to the hospital and perhaps a wait to see a medical pro-fessional. In this case, for example, approximately 25 minutes elapsed between the time the police stopped McNeely and the time his blood was drawn. App. 36, 38.
As noted, the fact that alcohol dissipates gradually from the bloodstream does not diminish the compelling need for a search – critical evidence is still disappearing. But the fact that the dissipation persists for some time means that the police – although they may not be able to do anything about it right away – may still be able to respond to the ongoing destruction of evidence later on.
There might, therefore, be time to obtain a warrant in many cases. As the Court explains, police can often request warrants rather quickly these days. At least 30 States provide for electronic warrant applications. See ante, at 10–12, and n. 4. In many States, a police officer can call a judge, convey the necessary information, and be authorized to affix the judge’s signature to a warrant. See, e.g., Ala. Rule Crim. Proc. 3.8(b) (2012–2013); Alaska Stat. §12.35.015 (2012); Idaho Code §§19–4404, 19–4406 (Lexis 2004); Minn. Rules Crim. Proc. 36.01–36.08 (2010 and Supp. 2013); Mont. Code Ann. §46–5–222 (2012); see generally NHTSA, Use of Warrants for Breath Test Refusal: Case Studies 6–32 (No. 810852, Oct. 2007) (overview of procedures in Arizona, Michigan, Oregon, and Utah). Utah has an e-warrant procedure where a police officer enters information into a system, the system notifies a prosecutor, and upon approval the officer forwards the information to a magistrate, who can electronically re-turn a warrant to the officer. Utah, e-Warrants: Cross Boundary Collaboration 1 (2008). Judges have been known to issue warrants in as little as five minutes. Bergreen, Faster Warrant System Hailed, Salt Lake Tribune, Dec. 26, 2008, p. B1, col. 1. And in one county in Kansas, police officers can e-mail warrant requests to judges’ iPads; judges have signed such warrants and e-mailed them back to officers in less than 15 minutes. Benefiel, DUI Search Warrants: Prosecuting DUI Refusals, 9 Kansas Prosecutor 17, 18 (Spring 2012). The police are presumably familiar with the mechanics and time involved in the warrant process in their particular jurisdiction.
In a case such as this, applying the exigent circumstances exception to the general warrant requirement of the Fourth Amendment seems straightforward: If there is time to secure a warrant before blood can be drawn, the police must seek one. If an officer could reasonably conclude that there is not sufficient time to seek and receive a warrant, or he applies for one but does not receive a response before blood can be drawn, a warrantless blood draw may ensue. See Tyler, supra, at 509; see also Illinois v Rodriguez, 497 U. S. 177–186 (1990) (“in order to satisfy the ‘reasonableness’ requirement of the Fourth Amendment, what is generally demanded of the many factual determinations that must regularly be made by .... police officer[s] conducting a search or seizure under one of the exceptions to the warrant requirement .... is not that they always be correct, but that they always be reasonable”); Terry v Ohio, 392 U. S. 1, 20 (1968) (“police must, whenever practicable, obtain advance judicial approval of searches and seizures through the warrant procedure”).
Requiring police to apply for a warrant if practicable increases the likelihood that a neutral, detached judicial officer will review the case, helping to ensure that there is probable cause for any search and that any search is reasonable. We have already held that forced blood draws can be constitutional – that such searches can be reasonable – but that does not change the fact that they are significant bodily intrusions. See Schmerber, 384 U. S., at 770 (upholding a warrantless forced blood draw but noting the “importance of informed, detached and deliberate determinations of the issue whether or not to invade another’s body in search of evidence of guilt” as “indisputable and great”). Requiring a warrant whenever practicable helps ensure that when blood draws occur, they are indeed justified.
At the same time, permitting the police to act without a warrant to prevent the imminent destruction of evidence is well established in Fourth Amendment law. There is no reason to preclude application of that exception in drunk driving cases simply because it may take the police some time to be able to respond to the undoubted destruction of evidence, or because the destruction occurs continuously over an uncertain period.
And that is so even in situations where police have requested a warrant but do not receive a timely response. An officer who reasonably concluded there was no time to secure a warrant may have blood drawn from a suspect upon arrival at a medical facility. There is no reason an officer should be in a worse position, simply because he sought a warrant prior to his arrival at the hospital.
The Court resists the foregoing, contending that the question presented somehow inhibits such a focused analysis in this case. See ante, at 20–23. It does not. The question presented is whether a warrantless blood draw is permissible under the Fourth Amendment “based upon the natural dissipation of alcohol in the bloodstream.” Pet. for Cert. i. The majority answers “It depends,” and so do I. The difference is that the majority offers no additional guidance, merely instructing courts and police officers to consider the totality of the circumstances. I believe more meaningful guidance can be provided about how to handle the typical cases, and nothing about the question presented prohibits affording that guidance.
A plurality of the Court also expresses concern that my approach will discourage state and local efforts to expedite the warrant application process. See ante, at 14. That is not plausible: Police and prosecutors need warrants in a wide variety of situations, and often need them quickly. They certainly would not prefer a slower process, just because that might obviate the need to ask for a warrant in the occasional drunk driving case in which a blood draw is necessary. The plurality’s suggestion also overlooks the interest of law enforcement in the protection a warrant provides.
The Court is correct when it says that every case must be considered on its particular facts. But the pertinent facts in drunk driving cases are often the same, and the police should know how to act in recurring factual situations. Simply put, when a drunk driving suspect fails field sobriety tests and refuses a breathalyzer, whether a warrant is required for a blood draw should come down to whether there is time to secure one.
Schmerber itself provides support for such an analysis. The Court there made much of the fact that “there was no time to seek out a magistrate and secure a warrant.” 384 U. S., at 771. It did so in an era when cell phones and e-mail were unknown. It follows quite naturally that if cell phones and e-mail mean that there is time to contact a magistrate and secure a warrant, that must be done. At the same time, there is no need to jettison the well-established exception for the imminent destruction of evidence, when the officers are in a position to do something about it.
* * *
Because the Missouri courts did not apply the rule I describe above, and because this Court should not do so in the first instance, I would vacate and remand for further proceedings in the Missouri courts.
This case requires the Court to decide whether the Fourth Amendment prohibits an officer from obtaining a blood sample without a warrant when there is probable cause to believe that a suspect has been driving under the influence of alcohol. Because the body’s natural metabolization of alcohol inevitably destroys evidence of the crime, it constitutes an exigent circumstance. As a result, I would hold that a warrantless blood draw does not violate the Fourth Amendment.
The Fourth Amendment states that “[t]he right of the people to be secure in their persons .... against unreasonable searches and seizures, shall not be violated, and no Warrants shall issue, but upon probable cause.” Before a search occurs, “a warrant must generally be secured,” Kentucky v King, 563 U. S. ___, ___ (2011) (slip op., at 5), but “this presumption may be overcome in some circumstances because ‘[t]he ultimate touchstone of the Fourth Amendment is “reasonableness.” ’” Ibid. (quoting Brig ham City v Stuart, 547 U. S. 398, 403 (2006); alteration in original).
The presence of “exigent circumstances” is one such exception to the warrant requirement. Exigency applies when “‘the needs of law enforcement [are] so compelling that [a] warrantless search is objectively reasonable under the Fourth Amendment.’” 563 U. S., at ___ (slip op., at 6) (quoting Mincey v Arizona, 437 U. S. 385, 394 (1978); second alteration in original). Thus, when exigent circumstances are present, officers may take actions that would typically require a warrant, such as entering a home in hot pursuit of a fleeing suspect. 563 U. S., at ___ (slip op., at 6). As relevant in this case, officers may also conduct a warrantless search when they have probable cause to believe that failure to act would result in “‘imminent destruction of evidence.’” Ibid. (quoting Brigham City, supra, at 403).
Once police arrest a suspect for drunk driving, each passing minute eliminates probative evidence of the crime. The human liver eliminates alcohol from the bloodstream at a rate of approximately 0.015 percent to 0.020 percent per hour, ante, at 8, with some heavy drinkers as high as 0.022 percent per hour, Brief for Petitioner 21 (citing medical studies), depending on, among other things, a person’s sex, weight, body type, and drinking history. Ante, at 8–9; Brief for United States as Amicus Curiae 23. The Court has acknowledged this fact since Schmerber v California, 384 U. S. 757, 770 (1966) (“We are told that the percentage of alcohol in the blood begins to diminish shortly after drinking stops, as the body functions to eliminate it from the system”). In that case, the Court recognized that destruction of evidence is inherent in drunk-driving cases and held that an officer investigating a drunk-driving crime “might reasonably [believe] that he [is] confronted with an emergency, in which the delay necessary to obtain a warrant, under the circumstances, threaten[s] ‘the destruction of evidence.’” Ibid. (quoting Preston v United States, 376 U. S. 364, 367 (1964)). The Court explained that drawing a person’s blood is “a highly effective means of determining the degree to which [he] is under the influence of alcohol” and is a reasonable procedure because blood tests are “commonplace” and “involv[e] virtually no risk, trauma, or pain.” 384 U. S., at 771. The Court, therefore, held that dissipation of alcohol in the blood constitutes an exigency that allows a blood draw without a warrant.
The rapid destruction of evidence acknowledged by the parties, the majority, and Schmerber’s exigency determination occurs in every situation where police have probable cause to arrest a drunk driver. In turn, that destruction of evidence implicates the exigent-circumstances doctrine. See Cupp v Murphy, 412 U. S. 291 (1973). In Cupp, officers questioning a murder suspect observed a spot on the suspect’s finger that they believed might be dried blood. Id., at 292. After the suspect began making obvious efforts to remove the spots from his hands, the officers took samples without obtaining either his consent or a warrant. Id., at 296. Following a Fourth Amendment challenge to this search, the Court held that the “ready destructibility of the evidence” and the suspect’s observed efforts to destroy it “justified the police in subjecting him to the very limited search necessary to preserve the highly evanescent evidence they found under his fingernails.” Ibid.
In this case, a similar exigency is present. Just as the suspect’s efforts to destroy “highly evanescent evidence” gave rise to the exigency in Cupp, the natural metabolization of blood alcohol concentration (BAC) creates an exigency once police have probable cause to believe the driver is drunk. It naturally follows that police may conduct a search in these circumstances.
A hypothetical involving classic exigent circumstances further illustrates the point. Officers are watching a warehouse and observe a worker carrying bundles from the warehouse to a large bonfire and throwing them into the blaze. The officers have probable cause to believe the bundles contain marijuana. Because there is only one person carrying the bundles, the officers believe it will take hours to completely destroy the drugs. During that time the officers likely could obtain a warrant. But it is clear that the officers need not sit idly by and watch the destruction of evidence while they wait for a warrant. The fact that it will take time for the evidence to be destroyed and that some evidence may remain by the time the officers secure a warrant are not relevant to the exigency. However, the ever-diminishing quantity of drugs may have an impact on the severity of the crime and the length of the sentence. See, e.g., 21 U. S. C. §841(b)(1)(D) (lower penalties for less than 50 kilograms of marijuana); United States Sentencing Commission, Guidelines Manual §2D1.1(c) (Nov. 2012) (drug quantity table tying base offence level to drug amounts). Conducting a warrantless search of the warehouse in this situation would be entirely reasonable.
The same obtains in the drunk-driving context. Just because it will take time for the evidence to be completely destroyed does not mean there is no exigency. Congress has conditioned federal highway grants on states’ adoption of laws penalizing the operation of a motor vehicle “with a blood alcohol concentration of 0.08 percent or greater.” 23 U. S. C. §163(a). See also 23 CFR §1225.1 (2012). All 50 States have acceded to this condition. National Highway Traffic Safety Admin. (NHTSA), Alcohol and Highway Safety: A Review of the State of Knowledge 167 (No. 811374, Mar. 2011) (NHTSA State Review); Mo. Ann. Stat. §§577.012(1)–(2) (West 2011) (establishing Missouri’s 0.08 percent BAC standard). Moreover, as of 2005, 32 States and the District of Columbia imposed additional penalties for BAC levels of 0.15 percent or higher. NHTSA State Review 175. Missouri is one such State. See, e.g., Mo. Stat. Ann. §§577.010(3)–(4), 577.012(4)–(5) (suspended sentence unavailable even for first offenders with BAC above 0.15 percent unless they complete drug treatment; mandatory jail time if treatment is not completed). As a result, the level of intoxication directly bears on enforcement of these laws. Nothing in the Fourth Amendment requires officers to allow evidence essential to enforcement of drunk-driving laws to be destroyed while they wait for a warrant to issue.
In today’s decision, the Court elides the certainty of evidence destruction in drunk-driving cases and focuses primarily on the time necessary for destruction. In doing so, it turns the exigency inquiry into a question about the amount of evidentiary destruction police must permit before they may act without a warrant. That inquiry is inconsistent with the actual exigency at issue: the uncontested destruction of evidence due to metabolization of alcohol. See Part I, supra. Moreover, the Court’s facts-and-circumstances analysis will be difficult to administer, a particularly important concern in the Fourth Amendment context.
The Court’s judgment reflects nothing more than a vague notion that everything will come out right most of the time so long as the delay is not too lengthy. Ante, at 12 (justifying delays in part because “BAC evidence is lost gradually and relatively predictably”); ante, at 10 (same, quoting Brief for Petitioner 27). But hard percentage lines have meaningful legal consequences in the drunk-driving context. The fact that police will be able to retrieve some evidence before it is all destroyed is simply not relevant to the exigency inquiry.
The majority believes that, absent special facts and circumstances, some destruction of evidence is acceptable. See ante, at 9 (“sufficient for our purposes to note that .... significant delay in testing will negatively affect the probative value” (emphasis added)). This belief must rest on the assumption that whatever evidence remains once a warrant is obtained will be sufficient to prosecute the suspect. But that assumption is clearly wrong. Suspects’ initial levels of intoxication and the time necessary to obtain warranted blood draws will vary widely from case to case. Even a slight delay may significantly affect probative value in borderline cases of suspects who are moderately intoxicated or suspects whose BAC is near a statutory threshold that triggers a more serious offence. See supra, at 4–5 (discussing laws penalizing heightened BAC levels). Similarly, the time to obtain a warrant can be expected to vary, and there is no reason to believe it will do so in a predictable fashion.
Further, the Court nowhere explains how an officer in the field is to apply the facts-and-circumstances test it adopts. First, officers do not have the facts needed to assess how much time can pass before too little evidence remains. They will never know how intoxicated a suspect is at the time of arrest. Otherwise, there would be no need for testing. Second, they will not know how long it will take to roust a magistrate from his bed, reach the hospital, or obtain a blood sample once there. As the Minnesota Supreme Court recognized in rejecting arguments like those adopted by the Court today:
[T]he officer has no control over how long it would take to travel to a judge or the judge’s availability. The officer also may not know the time of the suspect’s last drink, the amount of alcohol consumed, or the rate at which the suspect will metabolize alcohol. Finally, an officer cannot know how long it will take to obtain the blood sample once the suspect is brought to the hospital. Under a totality of the circumstances test, an officer would be called upon to speculate on each of these considerations and predict how long the most probative evidence of the defendant’s blood-alcohol level would continue to exist before a blood sample was no longer reliable.
State v Shriner, 751 N. W. 2d 538, 549 (2008) (footnote omitted).
The Court should not adopt a rule that requires police to guess whether they will be able to obtain a warrant before “too much” evidence is destroyed, for the police lack reliable information concerning the relevant variables.
This case demonstrates the uncertainty officers face with regard to the delay caused by obtaining a warrant. The arresting officer clearly had probable cause to believe respondent was drunk, but there was no way for the officer to quantify the level of intoxication to determine how quickly he needed to act in order to obtain probative evidence. Another officer testified at respondent’s trial that it typically took 1½ to 2 hours to obtain a drunk-driving warrant at night in Cape Girardeau County, Missouri. See App. 53–54. Respondent submitted an exhibit summarizing six late afternoon and nighttime drunk-driving search warrants that suggests the time may be shorter. Brief for Respondent 56; App. 70. Ultimately this factual tiff is beside the point; the spotty evidence regarding timing itself illustrates the fact that delays in obtaining warrants are unpredictable and potentially lengthy. A rule that requires officers (and ultimately courts) to balance transportation delays, hospital availability, and access to magistrates is not a workable rule for cases where natural processes inevitably destroy the evidence with every passing minute.
The availability of telephonic warrant applications is not an answer to this conundrum. See ante, at 10–12, and n. 4. For one thing, Missouri still requires written warrant applications and affidavits, Mo. Ann. Stat. §§542.276.2(1), 542.276.2.3 (West Supp. 2012), rendering the Court’s 50-State survey irrelevant to the actual disposition of this case. Ante, at 11, n. 4. But even if telephonic applications were available in Missouri, the same difficulties would arise. As the majority correctly recognizes, “[w]arrants inevitably take some time for police officers or prosecutors to complete and for magistrate judges to review.” Ante, at 12. During that time, evidence is destroyed, and police who have probable cause to believe a crime has been committed should not have to guess how long it will take to secure a warrant.
* * *
For the foregoing reasons, I respectfully dissent.
 As a result of his two prior drunk-driving convictions, McNeely was charged with a class D felony under Missouri law, which carries a maximum imprisonment term of four years. See Mo. Ann. Stat. §§558.011, 577.023.1(5), 577.023.3 (West 2011).
 Compare 358 S. W. 3d 65 (2012) (case below), State v Johnson, 744 N. W. 2d 340 (Iowa 2008) (same conclusion), and State v Rodriguez, 2007 UT 15, 156 P. 3d 771 (same), with State v Shriner, 751 N. W. 2d 538 (Minn. 2008) (holding that the natural dissipation of blood-alcohol evidence alone constitutes a per se exigency), State v Bohling, 173 Wis. 2d 529, 494 N. W. 2d 399 (1993) (same); State v Woolery, 116 Idaho 368, 775 P. 2d 1210 (1989) (same).
 We have recognized a limited class of traditional exceptions to the warrant requirement that apply categorically and thus do not require an assessment of whether the policy justifications underlying the exception, which may include exigency-based considerations, are im-plicated in a particular case. See, e.g., California v Acevedo, 500U. S. 565, 569–570 (1991) (automobile exception); United States v Robinson, 414 U. S. 218–235 (1973) (searches of a person incident to a lawful arrest). By contrast, the general exigency exception, which asks whether an emergency existed that justified a warrantless search, naturally calls for a case-specific inquiry.
 See Ala. Rule Crim. Proc. 3.8(b) (2012–2013); Alaska Stat. §12.35.015 (2012); Ariz. Rev. Stat. Ann. §§13–3914(C), 13–3915(D), (E) (West 2010); Ark. Code Ann. §16–82–201 (2005); Cal. Penal Code Ann. §1526(b) (West 2011); Colo. Rule Crim. Proc. 41(c)(3) (2012); Ga. Code Ann. §17–5–21.1 (2008); Haw. Rules Penal Proc. 41(h)–(i) (2013); Idaho Code §§19–4404, 19–4406 (Lexis 2004); Ind. Code §35–33–5–8 (2012); Iowa Code §§321J.10(3), 462A.14D(3) (2009) (limited to specific circumstances involving accidents); Kan. Stat. Ann. §§22–2502(a), 22–2504 (2011 Cum. Supp.); La. Code Crim. Proc. Ann., Arts. 162.1(B), (D) (West 2003); Mich. Comp. Laws Ann. §§780.651(2)–(6) (West 2006); Minn. Rules Crim. Proc. 33.05, 36.01–36.08 (2010 and Supp. 2013); Mont. Code Ann. §§46–5–221, 46–5–222 (2012); Neb. Rev. Stat. §§29–814.01, 29–814.03, 29–814.05 (2008); Nev. Rev. Stat. §§179.045(2), (4) (2011); N. H. Rev. Stat. Ann. §595–A:4–a (Lexis Supp. 2012); N. J. Rule Crim. Proc. 3:5–3(b) (2013); N. M. Rules Crim. Proc. 5–211(F)(3), (G)(3) (Supp. 2012); N. Y. Crim. Proc. Law Ann. §§690.35(1), 690.36(1), 690.40(3), 690.45(1), (2) (West 2009); N. C. Gen. Stat. Ann. §15A–245(a)(3) (Lexis 2011); N. D. Rules Crim. Proc. 41(c)(2)–(3) (2012–2013); Ohio Rules Crim. Proc. 41(C)(1)–(2) (2011); Okla. Stat. Ann., Tit. 22, §§1223.1, 1225(B) (West 2011); Ore. Rev. Stat. §§133.545(5)–(6) (2011); Pa. Rules Crim. Proc. 203(A), (C) (2012); S. D. Codified Laws §§23A–35–4.2, 23A–35–5, 23A–35–6 (2004); Utah Rule Crim. Proc. 40(l) (2012); Vt. Rules Crim. Proc. 41(c)(4), (g)(2) (Supp. 2012); Va. Code Ann. §19.2–54 (Lexis Supp. 2012); Wash. Super. Ct. Crim. Rule 2.3(c) (2002); Wis. Stat. §968.12(3) (2007–2008); Wyo. Stat. Ann. §31–6–102(d) (2011); see generally 2 W. LaFave, Search and Seizure §4.3(b), pp. 511–516, andn. 29 (4th ed. 2004) (describing oral search warrants and collecting state laws). Missouri requires that search warrants be in writing and does not permit oral testimony, thus excluding telephonic warrants. Mo. Ann. Stat. §§542.276.2(1), 542.276.3 (West Supp. 2012). State law does permit the submission of warrant applications “by facsimile or other electronic means.” §542.276.3.
 During the suppression hearing in this case, McNeely entered into evidence a search-warrant form used in drunk-driving cases by the prosecutor’s office in Cape Girardeau County, where the arrest took place. App. 61–69. The arresting officer acknowledged that he had used such forms in the past and that they were “readily available.” Id., at 41–42.
 The dissent claims that a “50-state survey [is] irrelevant to the actual disposition of this case” because Missouri requires written warrant applications. Post, at 8. But the per se exigency rule that the State seeks and the dissent embraces would apply nationally because it treats “the body’s natural metabolization of alcohol” as a sufficient basis for a warrantless search everywhere and always. Post, at 1. The technological innovations in warrant procedures that many Stateshave adopted are accordingly relevant to show that the per se rule is overbroad.
 The dissent contends that officers in the field will be unable to apply the traditional totality of the circumstances test in this context because they will not know all of the relevant facts at the time of an arrest. See post, at 6. But because “[t]he police are presumably familiar with the mechanics and time involved in the warrant process in their particular jurisdiction,” post, at 8 (opinion of Roberts, C. J.), we expect that officers can make reasonable judgments about whether the warrant process would produce unacceptable delay under the circumstances. Reviewing courts in turn should assess those judgments “‘from the perspective of a reasonable officer on the scene, rather than with the 20/20 vision of hindsight.’” Ryburn v Huff, 565 U. S. ___, ___ (2012) (per curiam) (slip op., at 8).
 Pursuant to congressional directive, the NHTSA conditions federal highway grants on States’ adoption of laws making it a per se offence to operate a motor vehicle with a BAC of 0.08 percent or greater. See 23 U. S. C. §163(a); 23 CFR §1225.1 (2012). Several federal prohibitions on drunk driving also rely on the 0.08 percent standard. E.g., 32 CFR §§234.17(c)(1)(ii), 1903.4(b)(1)(i)–(ii); 36 CFR §4.23(a)(2). In addition, 32 States and the District of Columbia have adopted laws that impose heightened penalties for operating a motor vehicle at or above a BAC of 0.15 percent. See NHTSA Review 175.
 See Ala. Code §32–5–192(c) (2010); Alaska Stat. §§28.35.032(a), 28.35.035(a) (2012); Ariz. Rev. Stat. Ann. §28–1321(D)(1) (West 2012); Ark. Code Ann. §§5–65–205(a)(1), 5–65–208(a)(1) (Supp. 2011);Conn. Gen. Stat. §§14–227b(b), 14–227c(b) (2011); Fla. Stat. Ann. §316.1933(1)(a) (West 2006); Ga. Code Ann. §§40–5–67.1(d), (d.1) (2011); Haw. Rev. Stat. §291E–15 (2009 Cum. Supp.), §§291E–21(a), 291E–33 (2007), §291E–65 (2009 Cum. Supp.); Iowa Code §§321J.9(1), 321J.10(1), 321J.10A(1) (2009); Kan. Stat. Ann. §§8–1001(b), (d) (2001); Ky. Rev. Stat. Ann. §189A.105(2) (Lexis Supp. 2012); La. Rev. Stat. Ann. §§32:666.A(1)(a)(i), (2) (Supp. 2013); Md. Transp. Code Ann. §§16–205.1(b)(i)(1), (c)(1) (Lexis 2012); Mass. Gen. Laws Ann., ch. 90, §§24(1)(e), (f)(1) (West 2012); Mich. Comp. Laws Ann. §257.625d(1) (West 2006); Miss. Code Ann. §63–11–21 (1973–2004); Mont. Code Ann. §§61–8–402(4), (5) (2011); Neb. Rev. Stat. §60–498.01(2) (2012Cum. Supp.), §60–6,210 (2010); N. H. Rev. Stat. Ann. §§265–A:14(I),265–A:16 (West 2012 Cum. Supp.); N. M. Stat. Ann. §66–8–111(A) (LexisNexis 2009); N. Y. Veh. & Traf. Law Ann. §§1194(2)(b)(1), 1194(3) (West 2011); N. D. Cent. Code Ann. §39–20–01.1(1) (Lexis Supp. 2011), §39–20–04(1) (Lexis 2008); Okla. Stat., Tit. 47, §753 (West Supp. 2013); Ore. Rev. Stat. §813.100(2) (2011); 75 Pa. Cons. Stat. §1547(b)(1) (2004); R. I. Gen. Laws §§31–27–2.1(b), 31–27–2.9(a) (Lexis 2010); S. C. Code Ann. §56–5–2950(B) (Supp. 2011); Tenn. Code Ann. §§55–10–406(a)(4), (f) (2012); Tex. Transp. Code Ann. §§724.012(b), 724.013 (West 2011); Vt. Stat. Ann., Tit. 23, §§1202(b), (f) (2007); Wash. Rev. Code §§46.20.308 (2)–(3), (5) (2012); W. Va. Code Ann. §17C–5–7 (Lexis Supp. 2012); Wyo. Stat. Ann. §31–6–102(d) (Lexis 2011).
 See Ariz. Rev. Stat. Ann. §28–1321(D)(1) (West 2012); Ga. Code Ann. §§40–5–67.1(d), (d.1) (2011); Ky. Rev. Stat. Ann. §189A.105(2)(b) (Lexis Supp. 2012); Mich. Comp. Laws Ann. §257.625d(1) (West 2006); Mont. Code Ann. §61–8–402(5) (2011); N. M. Stat. Ann. §66–8–111(A) (LexisNexis 2009); N. Y. Veh. & Traf. Law Ann. §§1194(2)(b)(1), 1194(3) (West 2011); Ore. Rev. Stat. 813.320(2)(b) (2011); R. I. Gen. Laws §31–27–2.9(a) (Lexis 2010); Tenn. Code Ann. §55–10–406(a)(4) (2012); Vt. Stat. Ann., Tit. 23, §1202(f) (2007); Wash. Rev. Code §46.20.308(1) (2012); W. Va. Code Ann. §17C–5–7 (Supp. 2012) (as interpreted in State v Stone, 229 W. Va. 271, ___, 728 S. E. 2d 155, 167–168 (2012)); Wyo. Stat. Ann. §31–6–102(d) (2011); see also State v Harris, 763 N. W. 2d 269, 273–274 (Iowa 2009) (per curiam) (recognizing that Iowa law imposes a warrant requirement subject to a limited case-specific exigency exception).
 No findings were made by the trial court concerning how long a warrant would likely have taken to issue under the circumstances. The minimal evidence presented on this point was not uniform. A second patrol officer testified that in a typical DWI case, it takes between 90 minutes and 2 hours to obtain a search warrant following an arrest. App. 53–54. McNeely, however, also introduced an exhibit documenting six recent search warrant applications for blood testing in Cape Girardeau County that had shorter processing times. Id., at 70.
 And that second-best evidence may prove useless. When experts have worked backwards to identify a defendant’s BAC at the time he was driving, defence attorneys have objected to that evidence, courts have at times rejected it, and juries may be suspicious of it. See, e.g., 1 D. Nichols & F. Whited, Drinking/Driving Litigation §2:9, pp. 2–130 to 2–137 (2d ed. 2006) (noting counsel objections to such evidence); State v Eighth Judicial District Court, 127 Nev. ___, 267 P. 3d 777 (2011) (affirming rejection of such evidence); L. Taylor & S. Oberman, Drunk Driving Defence §6.03 (7th ed. 2010) (describing ways to undermine such evidence before a jury).
 This case involves medical personnel drawing blood at a medical facility, not police officers doing so by the side of the road. See Schmerber v California, 384 U. S. 757–772 (1966) (“Petitioner’s blood was taken by a physician in a hospital environment according to accepted medical practices. We are thus not presented with the serious questions which would arise if a search involving use of a medical technique, even of the most rudimentary sort, were made by other than medical personnel or in other than a medical environment – for example, if it were administered by police in the privacy of the stationhouse”); Brief for Respondent 53, and n. 21 (describing roadside blood draws in Arizona). A plurality of the Court suggests that my approach could make roadside blood draws a more attractive option for police, but such a procedure would pose practical difficulties and, as the Court noted in Schmerber, would raise additional and serious Fourth Amendment concerns. See ante, at 14–15.
 Neither party has challenged this determination, which this Court has reaffirmed several times. See, e.g., Skinner v Railway Labor Executives’ Assn., 489 U. S. 602, 625 (1989); Winston v Lee, 470 U. S. 753–763 (1985).
 Because the Court’s position is likely to result in delay in obtaining BAC evidence, it also increases the likelihood that prosecutors will be forced to estimate the amount of alcohol in a defendant’s bloodstream using BAC numbers obtained hours later. In practice, this backwards extrapolation is likely to devolve into a battle of the experts, as each side seeks to show that stale evidence supports its position. There is no need for this outcome. Police facing inevitable destruction situations need not forgo collecting the most accurate available evidence simply because they might be able to use an expert witness and less persuasive evidence to approximate what they lost.
all rights reserved