dissenting.
Almost 10 years ago in Terry v. Ohio, 392 U. S. 1, the Court held that “probable cause” was not required to justify every seizure of the person by a police officer. That case was decided after six months of deliberation following full argument and unusually elaborate briefing.1 The approval in Terry of a lesser standard for certain limited situations represented a major development in Fourth Amendment jurisprudence.
Today, without argument, the Court adopts still another— *116and even lesser — standard of justification for a major category of police seizures.2 More importantly, it appears to abandon “the central teaching of this Court’s Fourth Amendment jurisprudence” 3 — which has ordinarily required individualized inquiry into the particular facts justifying every police intrusion — in favor of a general rule covering countless situations. But what is most disturbing is the fact that this important innovation is announced almost casually, in the course of explaining the summary reversal of a decision the Court should not even bother to review.
Since Mimms has already served his sentence, the importance of reinstating his conviction is minimal at best.4 Even if the Pennsylvania Supreme Court has afforded him greater protection than is required by the Federal Constitution, the conviction may be invalid under state law.5 Moreover, the *117Pennsylvania Supreme Court may still construe its own constitution to prohibit what it described as the “indiscriminate procedure” of ordering all traffic offenders out of their vehicles. 471 Pa. 546, 553, 370 A. 2d 1157, 1161.6 In all events, whatever error the state court has committed affects only the Commonwealth of Pennsylvania. Its decision creates no conflict requiring resolution by this Court on a national level. In most cases, these considerations would cause us to deny certiorari.
No doubt it is a legitimate concern about the safety of police officers throughout the Nation that prompts the Court to give this case such expeditious treatment. I share that concern and am acutely aware that almost every decision of this Court holding that an individual’s Fourth Amendment rights have been invaded makes law enforcement somewhat more difficult and hazardous. That, however, is not a sufficient reason for this Court to reach out to decide every new Fourth Amendment issue as promptly as possible. In this area of constitutional adjudication, as in all others, it is of paramount importance that the Court have the benefit of differing judicial evaluations of an issue before it is finally resolved on a nationwide basis.
This case illustrates two ways in which haste can introduce a new element of confusion into an already complex set of rules. First, the Court has based its legal ruling on a factual assumption about police safety that is dubious at best; second, the Court has created an entirely new legal standard of justification for intrusions on the liberty of the citizen.
Without any attempt to differentiate among the multitude of varying situations in which an officer may approach a person *118seated in an automobile, the Court characterizes the officer’s risk as “inordinate” on the basis of this statement:
“ 'According to one study, approximately 30% of police shootings occurred when a police officer approached a suspect seated in an automobile. Bristow, Police Officer Shootings — A Tactical Evaluation, 54 J. Crim. L. C. & P. S. 93 (1963).’ Adams v. Williams, 407 U. S. 143, 148 n. 3 (1972).” Ante, at 110.
That statement does not fairly characterize the study to which it refers. Moreover, the study does not indicate that police officers can minimize the risk of being shot by ordering drivers stopped for routine traffic violations out of their cars. The study reviewed 110 selected police shootings that occurred in 1959, 1960, and 1961.7 In 35 of those cases, “officers were attempting to investigate, control, or pursue suspects who were in automobiles.” 8 Within the group of 35 cases, there were examples of officers who “were shot through the windshield or car body while their vehicle was moving”; examples in which “the officer was shot while dismounting from his vehicle or while approaching the suspect[’]s vehicle”; and, apparently, instances in which the officer was shot by a passenger in the vehicle. Bristow, supra, n. 7, at 93.
In only 28 of the 35 cases was the location of the suspect who shot the officer verified. In 12 of those cases the suspect was seated behind the wheel of the car, but that figure seems to include cases in which the shooting occurred before the officer had an opportunity to order the suspect to get out. In *119nine cases the suspect was outside the car talking to the officer when the shooting occurred.
These figures tell us very little about the risk associated with the routine traffic stop; 9 and they lend no support to the Court’s assumption that ordering the routine traffic offender out of his car significantly enhances the officer’s safety. Arguably, such an order could actually aggravate the officer’s danger because the fear of a search might cause a serious offender to take desperate action that would be unnecessary if he remained in the vehicle while being ticketed. Whatever the reason, it is significant that some experts in this area of human behavior strongly recommend that the police officer “never allow the violator to get out of the car . . . .” 10
Obviously, it is not my purpose to express an opinion on the *120safest procedure to be followed in making traffic arrests or to imply that the arresting officer faces no significant hazard, even in the apparently routine situation. I do submit, however, that no matter how hard we try we cannot totally eliminate the danger associated with law enforcement, and that, before adopting a nationwide rule, we should give further consideration to the infinite variety of situations in which today’s holding may be applied.
The Court cannot seriously believe that the risk to the arresting officer is so universal that his safety is always a reasonable justification for ordering a driver out of his car. The commuter on his way home to dinner, the parent driving children to school, the tourist circling the Capitol, or the family on a Sunday aiternoon outing hardly pose the same threat as a driver curbed after a high-speed chase through a high-crime area late at night. Nor is it universally true that the driver’s interest in remaining in the car is negligible. A woman stopped at night may fear for her own safety; a person *121in poor health may object to standing in the cold or rain; another who left home in haste to drive children or spouse to school or to the train may not be fully dressed; an elderly driver who presents no possible threat of violence may regard the police command as nothing more than an arrogant and unnecessary display of authority. Whether viewed from the standpoint of the officer’s interest in his own safety, or of the citizen’s interest in not being required to obey an arbitrary command, it is perfectly obvious that the millions of traffic stops that occur every year are not fungible.
Until today the law applicable to seizures of a person has required individualized inquiry into the reason for each intrusion, or some comparable guarantee against arbitrary harassment.11 A factual demonstration of probable cause is required *122to justify an arrest; an articulable reason to suspect criminal activity and possible violence is needed to justify a stop and frisk. But to eliminate any requirement that an officer be able to explain the reasons for his actions signals an abandonment of effective judicial supervision of this kind of seizure and leaves police discretion utterly without limits. Some citizens will be subjected to this minor indignity while others— perhaps those with more expensive cars, or different bumper stickers, or different-colored skin — may escape it entirely.
The Court holds today that “third-class” seizures may be imposed without reason; how large this class of seizures may be or become we cannot yet know. Most narrowly, the Court has simply held that whenever an officer has an occasion to speak with the driver of a vehicle, he may also order the driver out of the car. Because the balance of convenience and danger is no different for passengers in stopped cars, the Court’s logic necessarily encompasses the passenger. This is true even though the passenger has committed no traffic offense. If the rule were limited to situations in which individualized inquiry identified a basis for concern in particular cases, then the character of the violation might justify different treatment of the driver and the passenger. But when the justification rests on nothing more than an assumption about the danger associated with every stop — no matter how trivial *123the offense — the new rule must apply to the passenger as well as to the driver.
If this new rule is truly predicated on a safety rationale— rather than a desire to permit pretextual searches — -it should also justify a frisk for weapons, or at least an order directing the driver to lean on the hood of the car with legs and arms spread out. For unless such precautionary measures are also taken, the added safety — if any — in having the driver out of the car is of no value when a truly dangerous offender happens to be caught.12
I am not yet persuaded that the interest in police safety requires the adoption of a standard any more lenient than that permitted by Terry v. Ohio.13 In this case the offense might well have gone undetected if respondent had not been ordered out of his car, but there is no reason to assume that he otherwise would have shot the officer. Indeed, there has been no showing of which I am aware that the Terry standard will not provide the police with a sufficient basis to take appropriate protective measures whenever there is any real basis for concern. When that concern does exist, they should be able to frisk a violator, but I question the need to eliminate the requirement of an articulable justification in each case and to authorize the indiscriminate invasion of the liberty of every citizen stopped for a traffic violation, no matter how petty.
Even if the Pennsylvania Supreme Court committed error, that is not a sufficient justification for the exercise of this *124Court’s discretionary power to grant review, or for the summary disposition of a novel constitutional question. For this kind of disposition gives rise to an unacceptable risk of error and creates “the unfortunate impression that the Court is more interested in upholding the power of the State than in vindicating individual rights.” Idaho Dept. of Employment v. Smith, ante, at 105 (Stevens, J., dissenting in part).
I respectfully dissent from the grant of certiorari and from the decision on the merits without full argument and briefing.
Briefs of amici curiae, urging reversal, were filed by Jack Greenberg, James M. Nabrit III, Michael Meltsner, Melvyn Zarr, and Anthony G. Amsterdam for the NAACP Legal Defense and Educational Fund, Inc., and by Bernard A. Berkman, Melvin L. Wulf, and Alan H. Levine for the American Civil Liberties Union et al.
Briefs of amici curiae, urging affirmance, were filed by Solicitor General Griswold, Assistant Attorney General Vinson, Ralph S. Spritzer, Beatrice Rosenberg, and Mervyn Hamburg for the United States; by Louis J. Lefkowitz, pro se, Samuel A. Hirshowitz, First Assistant Attorney General, and Maria L. Marcus and Brenda Soloff, Assistant Attorneys General, for the Attorney General of New York; by Charles Moylan, Jr., Bvelle J. Younger, and Harry Wood for the National District Attorneys’ Assn.; and by James R. Thompson for Americans for Effective Law Enforcement. See 392 U. S., at 4.
The Court does not dispute, nor do I, that ordering Mimms out of his ear was a seizure. A seizure occurs whenever an “officer, by means of physical force or show of authority, ... in some way restraints] the liberty of a citizen . . . .” Id., at 19 n. 16. See also Adams v. Williams, 407 U. S. 143, 146.
In Terry, the Court made it clear that the reasonableness of a search is to be determined by an inquiry into the facts of each case:
“[I]n justifying the particular intrusion the police officer must be able to point to specific and articulable facts which, taken together with rational inferences from those facts, reasonably warrant that intrusion.” 392 U. S., at 21.
In a footnote, the Court continued:
“This demand for specificity in the information upon which police action is predicated is the central teaching of this Court’s Fourth Amendment jurisprudence.” Id., at 21 n. 18 (citing a long list of authorities).
For the reasons stated in n. 3 of the Court’s opinion, I agree that the case is not moot. Nevertheless, the fact that the case barely escapes mootness supports the conclusion that certiorari should be denied.
Two members of the court were persuaded that introducing testimony about Mimms’ Muslim religious beliefs was prejudicial error, and three others specifically reserved the issue. 471 Pa. 546, 555 n. 2, and 556-557, 370 A. 2d 1157, 1158 n. 2, and 1162-1163.
Cf. State v. Opperman, 89 S. D. 25, 228 N. W. 2d 152 (1975), rev’d, 428 U. S. 364, judgment reinstated under state constitution,-S. D.-, 247 N. W. 2d 673 (1976).
As the author pointed out, “[n]o attempt was made to obtain a random selection of these cases, as they were extremely hard to collect.” Bristow, Police Officer Shootings — A Tactical Evaluation, 54 J. Crim. L. C. & P. S. 93 (1963).
Ibid. Since 35 is 32% of 110, presumably this is the basis for the “30%” figure used in the Court’s statement. As the text indicates, however, not all of these cases involved police officers approaching a 'parked vehicle. Whether any of the incidents involved routine traffic offenses, such as driving with an expired license tag, is not indicated in the study.
Over the past 10 years, more than 1,000 police officers have been murdered. FBI, Uniform Crime Reports 289 (1976). Approximately 10% of those killings, or about 11 each year, occurred during “traffic pursuits and stops,” but it is not clear how many of those pursuits and stops involved offenses such as reckless or high-speed driving, rather than offenses such as driving on an expired license, or how often the shootings could have been avoided by ordering the driver to dismount.
“2. Never allow the violator to get out of the car and stand to its left. If he does get out, which should be avoided, walk him to the rear and right side of the car. Quite obviously this is a much safer area to conduct a conversation.” V. Folley, Police Patrol Techniques and Tactics 95 (1973) (emphasis in original).
Another authority is even more explicit:
“The officer should stand slightly to the rear of the front door and doorpost. This will prevent the violator from suddenly opening the door and striking the officer. In order to thoroughly protect himself as much as possible, the officer should reach with his weak hand and push the lock button down if the window is open. This will give an indication to the driver that he is to remain inside the vehicle. It will also force the driver to turn his head to talk with the officer.
“The officer should advise the violator why he was stopped and then explain what action the officer intends to take, whether it is a verbal or written warning, or a written citation. If the suspect attempts to exit his vehicle, the officer should push the door closed, lock it, if possible, and *120tell the driver to ‘please stay in the car! ’ Then he should request [the] identification he desires and request the violator to hand the'material out of the window away from the vehicle. The officer should not stare at the identification but [should] return to his vehicle by backing away from the suspect car. As the patrolman backs away, he should keep his eyes on the occupant (s).
“The officer should remain outside of the patrol unit to use the radio or to write a ticket. The recommended position for him at this time would be to the right side of the patrol unit. Should the driver of the violator vehicle make exit from his seat, the officer should direct the violator to the rear center of his vehicle or the front center area of the patrol unit. Preferably, the officer should verbally attempt to get the violator co re-enter and remain in the vehicle.” A. Yount, Vehicle Stops Manual, Misdemeanor and Felony 2-3 (1976).
Conflicting advice is found in, an earlier work, G. Payton, Patrol Procedure 298 (4th ed. 1971). It is worth noting that these authorities suggest that any danger to the officer from passing traffic may be greatly reduced by the simple and unintrusive expedient of parking the police car behind, and two or three feet to the left of, the offender’s vehicle. Folley, supra, at 93; Payton, supra, at 301; Yount, supra, at 2.
Government instrusions must be justified with particularity in all but a few narrowly cabined contexts. Inspections pursuant to a general regulatory scheme and stops at border checkpoints are the best known exceptions to the particularity requirement. And even these limited exceptions fit within a broader rule — that the general populace should never be subjected to seizures without some assurance that the intruding officials are acting under a carefully limited grant of discretion. Health and safety inspections may be conducted only if the inspectors obtain warrants, though the warrants may be broader than the ordinary search warrant; officials may not wander at large in the city, conducting inspections without reason. Camara v. Municipal Court, 387 U. S. 523. Similar assurances of regularity and fairness can be found in public, fixed checkpoints:
“[Checkpoint operations both appear to and actually involve less discretionary enforcement activity [than stops by roving patrols]. The regularized manner in which established checkpoints are operated is visible evidence, reassuring to law-abiding motorists, that the stops are duly authorized and believed to serve the public interest. The location of a fixed checkpoint is not chosen by officers in the field, but by officials responsible for making overall decisions as to the most effective allocation of limited enforcement resources. We may assume that such officials will be unlikely to locate a checkpoint where it bears arbitrarily or oppressively on motorists as a class. And since field officers may stop only those cars passing the checkpoint, there is less room for abusive or harassing stops of *122individuals than ... in the case of roving-patrol stops.” United States v. Martinez-Fuerte, 428 U. S. 543, 559.
There is, of course, a general rule authorizing searches incident to full custodial arrests, but in such cases an individualized determination of probable cause adequately justifies both the search and the seizure. In that situation, unlike this one, the intrusion on the citizen’s liberty is “strictly circumscribed by the exigencies which justify its initiation.” Terry v. Ohio, 392 U. S. 1, 26. In this case, there was no custodial arrest, and I assume (perhaps somewhat naively) that the offense which gave rise to the stop of Mimms’ car would not have warranted a full custodial arrest without some additional justification. See Gustafson v. Florida, 414 U. S. 269, 266-267 (Stewart, J., concurring); id., at 238 n. 2 (Powell, J., concurring).
Terry v. Ohio, supra, at 33 (Harlan, J., concurring):
“Just as a full search incident to a lawful arrest requires no additional justification, a limited frisk incident to a lawful stop must often be rapid and routine. There is no reason why an officer, rightfully but forcibly confronting a person suspected of a serious crime, should have to ask one question and take the risk that the answer might be a bullet.”
do not foreclose the possibility that full argument would convince me that the Court’s analysis of the merits is correct. My limited experience has convinced me that one’s initial impression of a novel issue is frequently different from his final evaluation.