with whom Mr. Justice Douglas and Mr. Justice Brennan join, dissenting.
Certain fundamental principles have characterized this Court's Fourth Amendment jurisprudence over the years. Perhaps the most basic of these was expressed by Mr. Justice Butler, speaking for a unanimous Court in Go-Bart Co. v. United States, 282 U. S. 344 (1931): “There is no formula for the determination of reasonableness. Each case is to be decided on its own facts and circumstances.” Id., at 357. As we recently held: “The constitutional validity of a warrantless search is preeminently the sort of question which can only be decided in the concrete factual context of the individual case.” Sibron v. New York, 392 U. S. 40, 59 (1968). And the intensive, at times painstaking, case-by-case analysis characteristic of our Fourth Amendment decisions bespeaks our “jealous regard for maintaining the integrity of individual rights.” Mapp v. Ohio, 367 U. S. 643, 647 *239(1961). See also Weeks v. United States, 232 U. S. 383, 393 (1914).
In the present case, however, the majority turns its back on these principles, holding that “the fact of the lawful arrest” always establishes the authority to conduct a full search of the arrestee’s person, regardless of whether in a particular case “there was present one of the reasons supporting the authority for a search of the person incident to a lawful arrest.” Ante, at 235. The majority’s approach represents a clear and marked departure from our long tradition of case-by-case adjudication of the reasonableness of searches and seizures under the Fourth Amendment. I continue to believe that “[t]he scheme of the Fourth Amendment becomes meaningful only when it is assured that at some point the conduct of those charged with enforcing the laws can be subjected to the more detached, neutral scrutiny of a judge who must evaluate the reasonableness of a particular search or seizure in light of the particular circumstances.” Terry v. Ohio, 392 U. S. 1, 21 (1968). Because I find the majority’s reasoning to be at odds with these fundamental principles, I must respectfully dissent.
I
On April 19, 1968, Officer Richard Jenks stopped a 1965 Cadillac driven by respondent at the intersection of 9th and U Streets, N. W., in the District of Columbia, for what was called a “routine spot check.” At that time, Officer Jenks examined respondent’s temporary operator’s permit, automobile registration card, and Selective Service classification card. Although he permitted respondent to go on his way, Officer Jenks pursued a discrepancy he had noted between the “1938” date of birth given on the operator’s permit and the “1927” date of birth given on the Selective Service card. A check of police traffic records showed that an operator’s *240permit issued to one Willie Robinson, Jr., born in 1927, had been revoked, and that a temporary operator’s permit had subsequently been issued to one Willie Robinson, born in 1938. The pictures on the revoked permit and on the application for the temporary permit were of the same man — the person stopped by Jenks for the routine check on April 19. Having investigated the matter himself in this fashion, it is clear that Officer Jenks had probable cause to believe that respondent had violated a provision of the District of Columbia Motor Vehicle Code making it unlawful for any person to operate a motor vehicle in the District during the period for which his operator’s permit is revoked. D. C. Code Ann. § 40-302 (d) (1967).
Four days later, on April 23, 1968, while on duty in their patrol car, Officer Jenks and his partner saw respondent driving the same vehicle. They pulled up behind respondent’s car and signaled it to stop. From all indications in the record, respondent immediately complied and brought his car to a stop alongside the curb, the officers parking their patrol car immediately behind his.
Respondent got out of his car and walked back toward the patrol car. Both Officer Jenks and his partner got out of the patrol car and started toward respondent’s car. Officer Jenks asked respondent for his permit and registration card and, when shown the same permit respondent had given him four days earlier, informed respondent that he was under arrest for operating a motor vehicle after revocation of his operator’s permit.
Jenks then began to search respondent. His normal procedure in conducting a search of an arrestee would be to “have him spread-eagle over a wall or something of that nature.” But in Jenks’ own words, “I think almost every search is different. It depends on the man’s size and the nature of the crime.” Since he had a substantial *241height and weight advantage over respondent, and because the arrest was only for a traffic offense, Jenks chose instead to conduct the search face to face, in contrast to his normal practice.
The first step in the search was for Jenks to place both his hands on respondent's chest and begin to pat him down. During this patdown, Jenks felt something in the left breast pocket of respondent's heavy overcoat. Jenks later testified that he could not immediately tell what was in the pocket. The record does indicate, however, that the object did not feel like a gun and that Jenks had no particular indication it was a weapon of any kind.1 Nonetheless, he reached into the pocket and took the object out. It turned out to be a crumpled-up cigarette package.
With the package now in his hands, Jenks could feel objects inside but could not tell what they were. It does not appear that Jenks had any reason to believe, or did in fact believe, that the objects were weapons of any sort. He nevertheless opened up the package and looked inside, thereby finding the gelatin capsules of heroin which were introduced against respondent at his trial for the possession and facilitation of concealment of heroin.
H-i HH
Mr. Justice Jackson, writing for the Court in Johnson v. United States, 333 U. S. 10, 13-14 (1948), explained:
“The point of the Fourth Amendment, which often is not grasped by zealous officers, is not that it denies *242law enforcement the support of the usual inferences which reasonable men draw from evidence. Its protection consists in requiring that those inferences be drawn by a neutral and detached magistrate instead of being judged by the officer engaged in the often competitive enterprise of ferreting out crime. . . . When the right of privacy must reasonably yield to the right of search is, as a rule, to be decided by a judicial officer, not by a policeman or government enforcement agent.”
See also Coolidge v. New Hampshire, 403 U. S. 443, 449 (1971). The majority’s fear of overruling the “quick ad hoc judgment” of the police officer is thus inconsistent with the very function of the Amendment — to ensure that the quick ad hoc judgments of police officers are subject to review and control by the judiciary.
In the vast majority of cases, the determination of when the right of privacy must reasonably yield to the right of search is required to be made by a neutral judicial officer before the search is conducted. See Katz v. United States, 389 U. S. 347, 356-357 (1967); Agnello v. United States, 269 U. S. 20, 33 (1925). The Constitution requires that “the deliberate, impartial judgment of a judicial officer ... be interposed between the citizen and the police . . . .” Wong Sun v. United States, 371 U. S. 471, 481-482 (1963).
The requirement that the police seek prior approval of a search from a judicial officer is, no doubt, subject to “a few specifically established and well-delineated exceptions,” Katz v. United States, supra, at 357; including searches of a moving vehicle, Carroll v. United States, 267 U. S. 132 (1925); searches in certain exigent circumstances, Warden v. Hayden, 387 U. S. 294, 298-299 (1967); McDonald v. United States, 335 U. S. 451, 454-455 (1948); and searches incident to a lawful *243arrest, Agnello v. United States, supra; Chimel v. California, 395 U. S. 752 (1969). But because an exception is invoked to justify a search without a warrant does not preclude further judicial inquiry into the reasonableness of that search. It is the role of the judiciary, not of police officers, to delimit the scope of exceptions to the warrant requirement. “[T]he general requirement that a search warrant be obtained is not lightly to be dispensed with, and 'the burden is on those seeking [an] exemption [from the requirement] to show the need for it . . . ” Id., at 762, quoting United States v. Jeffers, 342 U. S. 48, 51 (1951). Exceptions to the warrant requirement are not talismans precluding further judicial inquiry whenever they are invoked, see Coolidge v. New Hampshire, supra, at 461, but rather are “jealously and carefully drawn.” Jones v. United States, 357 U. S. 493, 499 (1958).
Carrying out our mandate of delineating the proper scope of the search-incident-to-arrest exception requires consideration of the purposes of that exception as they apply to the particular search that occurred in this case. See Chimel v. California, supra, at 762-763; Preston v. United States, 376 U. S. 364, 367 (1964). Yet the majority, rather than focusing on the facts of this case, places great emphasis on the police department order which instructed Officer Jenks to conduct a full search and to examine carefully everything he found whenever making an in-custody arrest. See ante, at 221 and n. 2. But this mode of analysis was explicitly re'jected in Sibron v. New York, 392 U. S. 40 (1968). There both the defendant and the State urged that the principal issue before us was the constitutionality of a state statute which authorized the search there in question. We declined, however, to engage in what we viewed “as the abstract and unproductive exercise” of laying the words of the statute next to the Fourth Amendment “in an effort to determine whether the two are in some *244sense compatible.” Id., at 59. “Our constitutional inquiry,” we concluded, “would not be furthered here by an attempt to pronounce judgment on the words of the statute. We must confine our review instead to the reasonableness of the searches and seizures which underlie these two convictions.” Id., at 62.
The majority also suggests that the Court of Appeals reached a novel and unprecedented result by imposing qualifications on the historically recognized authority to conduct a full search incident to a lawful arrest. Nothing could be further from the truth, as the Court of Appeals itself was so careful to point out.
One need not go back to Blackstone’s Commentaries, Holmes’ Common Law, or Pollock & Maitland in search of precedent for the approach adopted by the Court of Appeals. Indeed, given the fact that mass production of the automobile did not begin until the early decades of the present century, I find it somewhat puzzling that the majority even looks to these sources for guidance on the only question presented in this case: the permissible scope of a search of the person incident to a lawful arrest for violation of a motor vehicle regulation. The fact is that this question has been considered by several state and federal courts, the vast majority of which have held that, absent special circumstances, a police officer has no right to conduct a full search of the person incident to a lawful arrest for violation of a motor vehicle regulation.
In Barnes v. State, 25 Wis. 2d 116, 130 N. W. 2d 264 (1964), for example, police officers stopped a car for a brake-light violation. Rather than simply issue a citation, the officers placed the driver under arrest. A full search of the driver’s person was then conducted, including shining a flashlight into his overcoat pocket, disclosing a small quantity of marihuana and a package of cigarette papers. The Supreme Court of Wisconsin held *245the search of the driver’s pocket unreasonable. While expressly holding that where a traffic offender is actually arrested, as distinguished from being given a summons, it is reasonable for the arresting officer to search his person for weapons, nevertheless the court held it unreasonable to look inside the driver’s overcoat pocket with a flashlight. “We cannot conceive,” the court said, “that this aspect of the search was a legitimate search for weapons. . . . We reject the state’s contention that any search of the person of one lawfully arrested is a valid search.” Id., at 126, 130 N. W. 2d, at 269.
In State v. Curtis, 290 Minn. 429, 190 N. W. 2d 631 (1971), police officers stopped a car which had defective taillights and which had made an illegal right turn. The officers decided to take the driver down to the station house and searched him for weapons before putting him in the squad car. One of the officers felt the outside of the driver’s pockets. As in Robinson’s case, the officer “detected some object but couldn’t tell what it was. It did not feel like a gun or knife.” Id., at 430, 190 N. W. 2d, at 632. “Neither officer expressed any concern for his personal safety. There was no testimony that they suspected defendant of any other criminal activity or were aware of any dangerous propensities on his part.” Id., at 431, 190 N. W. 2d, at 633. Nevertheless the officer reached into the pocket, resulting in the discovery of a package of marihuana. The Minnesota Supreme Court held the search unlawful. While recognizing the “concern for the injuries and loss of life experienced by police officers in face-to-face confrontations with traffic offenders,” the court held that “the validity of a search for weapons following a traffic arrest depends on whether the officer had reasonable grounds to believe a search was necessary for his own safety or to prevent an escape.” Id., at 436-437, 190 N. W. 2d, at 636, citing Shelton *246v. State, 3 Md. App. 394, 399, 239 A. 2d 610, 613 (1968).
Of like import is the decision of the Oregon Supreme Court in State v. O’Neal, 251 Ore. 163, 444 P. 2d 951 (1968) (en banc). Here defendant’s automobile was stopped because it had no rear license plate. When asked to produce an operator’s license, the defendant produced a temporary operator’s license issued to another person which had expired several years earlier. The officers then arrested defendant and placed him in the back seat of the police car. “One of the officers got in the police car and asked the defendant to remove his money from his wallet and give his wallet to the officer. The defendant did so and the officer took papers from the wallet and examined them. When the officer unfolded one piece of paper a half-smoked marijuana cigarette fell out.” Id., at 164-165, 444 P. 2d, at 952. The court held the search unlawful. Again, while recognizing the officer’s right to conduct a search incident to arrest in order to protect the officer and deprive the prisoner of potential means of escape, the court held:
“The search of the wallet obviously had nothing to do with the officers’ safety. The defendant testified that the officers 'patted him down’ before placing him in the police car. The officers did not remember whether they had or had not. In any event, it is difficult to see how defendant’s wallet could have reasonably been believed to have contained a weapon.” Id., at 166, 444 P. 2d, at 953.
See also People v. Marsh, 20 N. Y. 2d 98, 228 N. E. 2d 783 (1967); People v. Superior Court of Los Angeles County, 7 Cal. 3d 186, 496 P. 2d 1205 (1972); State v. Quintana, 92 Ariz. 267, 376 P. 2d 130 (1962) (en banc); People v. Zeigler, 358 Mich. 355, 100 N. W. 2d 456 (1960). The Tenth Circuit has likewise stated that it is “in complete agreement with the prevailing federal and state authority *247which condemns the search of persons and automobiles following routine traffic violations.” United States v. Humphrey, 409 F. 2d 1055, 1058 (1969). See also Amador-Gonzalez v. United States, 391 F. 2d 308, 315 (CA5 1968) (Wisdom, J.).
Accordingly, I think it disingenuous for the Court to now pronounce that what precedents exist on the question “tend to support the broad statement of the authority to search incident to arrest found in the successive decisions of this Court, rather than the restrictive one which was applied by the Court of Appeals in this case.” 2 Ante, at 232-233. It is disquieting, to say the least, to see the Court at once admit that “[virtually all of the statements of this Court affirming the existence of an unqualified authority to search incident to a lawful arrest are dicta” and concede that we are presented with an open question on which “further examination into history and practice” would be helpful, yet then conduct an examina*248tion into prior practice which is not only wholly superficial, but totally inaccurate and misleading.
The majority’s attempt to avoid case-by-case adjudication of Fourth Amendment issues is not only misguided as a matter of principle, but is also doomed to fail as a matter of practical application. As the majority itself is well aware, see ante, at 221 n. 1, the powers granted the police in this case are strong ones, subject to potential abuse. Although, in this particular case, Officer Jenks was required by police department regulations to make an in-custody arrest rather than to issue a citation, in most jurisdictions and for most traffic offenses the determination of whether to issue a citation or effect a full arrest is discretionary with the officer. There is always the possibility that a police officer, lacking probable cause to obtain a search warrant, will use a traffic arrest as a pretext to conduct a search. See, e. g., Amador-Gonzalez v. United States, supra. I suggest this possibility not to impugn the integrity of our police, but merely to point out that case-by-case adjudication will always be necessary to determine whether a full arrest was effected for purely legitimate reasons or, rather, as a pretext for searching the arrestee. “An arrest may not be used as a pretext to search for evidence.” United States v. Lefkowitz, 285 U. S. 452, 467 (1932). See also Jones v. United States, 357 U. S., at 500; Abel v. United States, 362 U. S. 217, 226 and 230 (1960); United States v. Rabinowitz, 339 U. S. 56, 82 (1950) (Frankfurter, J., dissenting). Cf. Chimel v. California, 395 U. S., at 767-768.
Ill
The majority states that “[a] police officer’s determination as to how and where to search the person of a suspect whom he has arrested is' necessarily a quick ad hoc judgment which the Fourth Amendment does not *249require to be broken down in each instance into an analysis of each step in the search.” 3 Ante, at 235. No precedent is cited for this broad assertion — not surprisingly, since there is none. Indeed, we only recently rejected such “a rigid all-or-nothing model of justification and regulation under the Amendment, [for] it obscures the utility of limitations upon the scope, as well as the initiation, of police action as a means of constitutional regulation. This Court has held in the past that a search which is reasonable at its inception may violate the Fourth Amendment by virtue of its intolerable intensity and scope.” Terry v. Ohio, 392 U. S., at 17-18. As we there concluded, “in determining whether the seizure and search were 'unreasonable’ our inquiry is a dual one — whether the officer’s action was justified at its inception, and whether it was reasonably related in scope to the circumstances which justified the interference in the first place.” Id., at 19-20.
As I view the matter, the search in this case divides into three distinct phases: the patdown of respondent’s coat pocket; the removal of the unknown object from the *250pocket; and the opening of the crumpled-up cigarette package.
A
No question is raised here concerning the lawfulness of the patdown of respondent’s coat pocket. The Court of Appeals unanimously affirmed the right of a police officer to conduct a limited frisk for weapons when making an in-custody arrest, regardless of the nature of the crime for which the arrest was made. As it said:
“[I]t would seem clearly unreasonable to expect a police officer to place a suspect in his squad car for transportation to the stationhouse without first taking reasonable measures to insure that the suspect is unarmed. We therefore conclude that whenever a police officer, acting within the bounds of his authority, makes an in-custody arrest, he may also conduct a limited frisk of the suspect’s outer clothing in order to remove any weapons the suspect may have in his possession.” 153 U. S. App. D. C. 114, 130, 471 F. 2d 1082, 1098 (1972) (footnote omitted; emphasis in original).
B
With respect to the removal of the unknown object from the coat pocket, the first issue presented is whether that aspect of the search can be sustained as part of the limited frisk for weapons. The weapons search approved by the Court of Appeals was modeled upon the narrowly drawn protective search for weapons authorized in Terry, which consists “of a limited patting of the outer clothing of the suspect for concealed objects which might be used as instruments of assault.” See Sibron v. New York, 392 U. S., at 65. See also Terry, supra, at 30.
It appears to have been conceded by the Government below that the removal of the object from respondent’s coat pocket exceeded the scope of a Terry frisk for *251weapons, since, under Terry, an officer may not remove an object from the suspect’s pockets unless he has reason to believe it to be a dangerous weapon. 153 U. S. App. D. C., at 121 and n. 9, 471 F. 2d, at 1089 and n. 9, citing ALI Model Code of Pre-Arraignment Procedure § 110.2 (4) (Proposed Official Draft No. 1, 1972). Cf. Sibron v. New York, supra, at 65.4
In the present case, however, Officer Jenks had no reason to believe and did not in fact believe that the object in respondent’s coat pocket was a weapon. He admitted later that the object did not feel like a gun. See n. 1, supra. In fact, he did not really have any thoughts one way or another about what was in the pocket. As Jenks himself testified, “I just searched him. I didn’t think about what I was looking for. I just searched him.” Since the removal of the object from the pocket cannot be justified as part of a limited Terry weapons frisk, the question arises whether it is reasonable for a police officer, when effecting an in-custody arrest of a traffic offender, to make a fuller search of the person than is permitted pursuant to Terry.
The underlying rationale of a search incident to arrest of a traffic offender initially suggests as reasonable a search whose scope is similar to the protective weapons frisk permitted in Terry. A search incident to arrest, as the majority indicates, has two basic functions: the removal of weapons the arrestee might use to resist arrest or effect an escape, and the seizure of evidence or fruits of the crime for which the arrest is made, so as to prevent their concealment or destruction. See ante, at 234; Chimel v. California, 395 U. S., at 763.
*252The Government does not now contend that the search of respondent's pocket can be justified by any need to find and seize evidence in order to prevent its concealment or destruction, for, as the Court of Appeals found, there is no evidence or fruits of the offense with which respondent was charged. The only rationale for a search in this case, then, is the removal of weapons which the arrestee might use to harm the officer and attempt an escape. This rationale, of course, is identical to the rationale of the search permitted in Terry. As we said there, “The sole justification of the search in the present situation is the protection of the police officer and others nearby, and it must therefore be confined in scope to an intrusion reasonably designed to discover guns, knives, clubs, or other hidden instruments for the assault of the police officer.” Terry v. Ohio, supra, at 29. Since the underlying rationale of a Terry search and the search of a traffic violator are identical, the Court of Appeals held that the scope of the searches must be the same. And in view of its conclusion that the removal of the object from respondent’s coat pocket exceeded the scope of a lawful Terry frisk, a conclusion not disputed by the Government or challenged by the majority here, the plurality of the Court of Appeals held that the removal of the package exceeded the scope of a lawful search incident to arrest of a traffic violator.
The problem with this approach, however, is that it ignores several significant differences between the context in which a search incident to arrest for a traffic violation is made, and the situation presented in Terry. Some of these differences would appear to suggest permitting a more thorough search in this case than was permitted in Terry; other differences suggest a narrower, more limited right to search than was there recognized.
The most obvious difference between the two contexts relates to whether the officer has cause to believe that *253the individual he is dealing with possesses weapons which might be used against him. Terry did not permit an officer to conduct a weapons frisk of anyone he lawfully stopped on the street, but rather, only where “he has reason to believe that he is dealing with an armed and dangerous individual . . . 392 U. S., at 27. While the policeman who arrests a suspected rapist or robber may well have reason to believe he is dealing with an armed and dangerous person, certainly this does not hold true with equal force with respect to a person arrested for a motor vehicle violation of the sort involved in this case.
Nor was there any particular reason in this case to believe that respondent was dangerous. He had not attempted to evade arrest, but had quickly complied with the police both in bringing his car to a stop after being signaled to do so and in producing the documents Officer Jenks requested. In fact, Jenks admitted that he searched respondent face to face rather than in spread-eagle fashion because he had no reason to believe respondent would be violent.
While this difference between the situation presented in Terry and the context presented in this case would tend to suggest a lesser authority to search here than was permitted in Terry, other distinctions between the two cases suggest just the opposite. As the Court of Appeals noted, a crucial feature distinguishing the in-custody arrest from the Terry context “ 'is not the greater likelihood that a person taken into custody is armed, but rather the increased likelihood of danger to the officer if in fact the person is armed.’ ” 153 U. S. App. D. C., at 130, 471 F. 2d, at 1098, quoting People v. Superior Court of Los Angeles County, 7 Cal. 3d, at 214, 496 P. 2d, at 1225 (Wright, C. J., concurring) (emphasis in original). A Terry stop involves a momentary encounter between officer and suspect, while an in-custody arrest places the two in close proximity *254for a much longer period of time. If the individual happens to have a weapon on his person, he will certainly have much more opportunity to use it against the officer in the in-custody situation. The prolonged proximity also makes it more likely that the individual will be able to extricate any small hidden weapon which might go undetected in a weapons frisk, such as a safety pin or razor blade. In addition, a suspect taken into custody may feel more threatened by the serious restraint on his liberty than a person who is simply stopped by an officer for questioning, and may therefore be more likely to resort to force.
Thus, in some senses there is less need for a weapons search in the in-custody traffic arrest situation than in a Terry context; while in other ways, there is a greater need. Balancing these competing considerations in order to determine what is a reasonable warrantless search in the traffic arrest context is a difficult process, one for which there may be no easy analytical guideposts. We are dealing with factors not easily quantified and,, therefore, not easily weighed one against the other. And the competing interests we are protecting — the individual’s interest in remaining free from unnecessarily intrusive invasions of privacy and society’s interest that police officers not take unnecessary risks in the performance of their duties — are each deserving of our most serious attention and do not themselves tip the balance in any particular direction.
As will be explained more fully below, I do not think it necessary to solve this balancing equation in this particular case. It is important to note, however, in view of the reasoning adopted by the majority, that available empirical evidence supports the result reached by the plurality of the Court of Appeals, rather than the result reached by the Court today.
*255The majority relies on statistics indicating that a significant percentage of murders of police officers occurs when the officers are making traffic stops. But these statistics only confirm what we recognized in Terry — that “American criminals have a long tradition of armed violence, and every year in this country many law enforcement officers are killed in the line of duty, and thousands more are wounded.” Terry v. Ohio, supra, at 23. As the very next sentence in Terry recognized, however, “[virtually all of these deaths and a substantial portion of the injuries are inflicted with guns and knives.” Id., at 24. The statistics relied on by the Government in this case support this observation. Virtually all of the killings are caused by guns and knives, the very type of weapons which will not go undetected in a properly conducted weapons frisk.5 It requires more than citation to these statistics, then, to support the proposition that it is reasonable for police officers to conduct more than a Terry-type frisk for weapons when seeking to disarm a traffic offender who is taken into custody.
C
The majority opinion fails to recognize that the search conducted by Officer Jenks did not merely involve a search of respondent’s person. It also included a separate search of effects found on his person. And even were we to assume, arguendo, that it was reasonable for Jenks to remove the object he felt in respondent’s pocket, clearly there was no justification consistent with *256the Fourth Amendment which would authorize his opening the package and looking inside.
To begin with, after Jenks had the cigarette package in his hands, there is no indication that he had reason to believe or did in fact believe that the package contained a weapon. More importantly, even if the crumpled-up cigarette package had in fact contained some sort of small weapon, it would have been impossible for respondent to have used it once the package was in the officer’s hands. Opening the package, therefore, did not further the protective purpose of the search. Even the dissenting opinion in the Court of Appeals conceded that “since the package was now in the officer’s possession, any risk of the prisoner’s use of a weapon in this package had been eliminated.” 6 153 U. S. App. D. C., at 150, 471 F. 2d, at 1118 (Wilkey, J., dissenting).
It is suggested, however, that since the custodial arrest itself represents a significant intrusion into the privacy of the person, any additional intrusion by way of opening or examining effects found on the person is not worthy of constitutional protection. But such an approach was expressly rejected by the Court in Chimel. There it *257was suggested that since the police had lawfully entered petitioner's house to effect an arrest, the additional invasion of privacy stemming from an accompanying search of the entire house was inconsequential. The Court answered: “[W]e can see no reason why, simply because some interference with an individual’s privacy and freedom of movement has lawfully taken place, further intrusions should automatically be allowed despite the absence of a warrant that the Fourth Amendment would otherwise require.” 395 U. S., at 766-767, n. 12.
The Fourth Amendment preserves the right of “the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures . . . .” Chimel established the principle that the lawful right of the police to interfere with the security of the person did not, standing alone, automatically confer the right to interfere with the security and privacy of his house. Hence, the mere fact of an arrest should be no justification, in and of itself, for invading the privacy of the individual’s personal effects.
The Government argues that it is difficult to see what constitutionally protected “expectation of privacy” a prisoner has in the interior of a cigarette pack. One wonders if the result in this case would have been the same were respondent a businessman who was lawfully taken into custody for driving without a license and whose wallet was taken from him by the police. Would it be reasonable for the police officer, because of the possibility that a razor blade was hidden somewhere in the wallet, to open it, remove all the contents, and examine each item carefully? Or suppose a lawyer lawfully arrested for a traffic offense is found to have a sealed envelope on his person. Would it be permissible for the arresting officer to tear open the envelope in order to make sure that it did not contain a clandestine weapon— perhaps a pin or a razor blade? Cf. Harris v. United *258States, 331 U. S. 145 (1947); Chimel v. California, supra, at 758. Would it not be more consonant with the purpose of the Fourth Amendment and the legitimate needs of the police to require the officer, if he has any question whatsoever about what the wallet or letter contains, to hold on to it until the arrestee is brought to the precinct station? 7
*259I, for one, cannot characterize any of these intrusions into the privacy of an individual’s papers and effects as being negligible incidents to the more serious intrusion into the individual’s privacy stemming from the arrest itself. Nor can any principled distinction be drawn between the hypothetical searches I have posed and the search of the cigarette package in this case. The only reasoned distinction is between warrantless searches which serve legitimate protective and evidentiary functions and those that do not. See Chimel, supra, at 766.
The search conducted by Officer Jenks in this case went far beyond what was reasonably necessary to protect him from harm or to ensure that respondent would not effect an escape from custody. In my view, it therefore fell outside the scope of a properly drawn “search incident to arrest” exception to the Fourth Amendment’s warrant requirement. I would affirm the judgment of the Court of Appeals holding that the fruits of the search should have been suppressed at respondent’s trial.
At the suppression hearing, Officer Jenks was shown a small derringer and was asked, after the gun was placed in a coat pocket, “whether or not it feels like the lump you felt in Mr. Robinson’s pocket the night you arrested him?” Jenks answered that the object “does not feel the same” but rather felt “harder,” “larger,” and “[m]uch more suspicious.”
Even the Court’s attempt to dip into the English common law is selective. The power to conduct a search incident to arrest was litigated in Leigh v. Cole, 6 Cox C. C. 329 (Oxford Circuit 1853), a civil case in which the plaintiff, a lawyer, was stopped while on the road and arrested by defendant superintendent of police. After the plaintiff was taken to the station house, a police constable searched him, at the defendant’s direction, and a tobacco box and a piece of paper were taken from him. In instructing the jury, the learned judge stated: “With respect to searching a prisoner, there is no doubt that a man when in custody may so conduct himself, by reason of violence of language or conduct, that a police officer may reasonably think it prudent and right to search him, in order to ascertain whether he has any weapon with which he might do mischief to the person or commit a breach of the peace; but at the same time it is quite wrong to suppose that a general rule can be applied to such a case. Even when a man is confined for being drunk and disorderly, it is not correct to say that he must submit to the degradation of being searched, as the searching of such a person must depend upon all the circumstances of the case.” Id., at 332.
The majority’s reference to the quick ad hoc judgment of the police officer may be read as an expression of doubt whether this kind of on-the-street police activity can effectively be controlled by court-imposed standards. This problem was discussed in Terry v. Ohio, 392 U. S. 1 (1968), where we recognized “the limitations of the judicial function in controlling the myriad daily situations in which policemen and citizens confront each other on the street.” Id., at 12. But as we concluded there, even though “[n]o judicial opinion can comprehend the protean variety of the street encounter, . . . courts still retain their traditional responsibility to guard against police conduct . . . which trenches upon personal security without the objective evidentiary justification which the Constitution requires. When such conduct is identified, it must be condemned by the judiciary and its fruits must be excluded from evidence in criminal trials.” Id., at 15.
This was also the position of the Police Department itself. Sergeant Donaldson, a Police Department Training Division instructor, testified: “If [the officer] could determine in his pat-down or frisk by squeezing that it was not, in fact, a weapon that could be used against him, then we don’t instruct him to go further.”
The Uniform Crime Reports prepared by the Federal Bureau of Investigation which are relied on by the majority, see ante, at 234 n. 5, indicate that 112 police officers were killed nationwide in 1972. Of these, 108 were killed by firearms. Two of the remaining four were killed with knives, and the last two cases involved a bomb and an automobile.
The dissent argued, however, that “further inspection of the package was still justifiable as a protective measure. If the package had contained a razor blade, or live bullets, the officer would have been alerted to search Robinson much more thoroughly since the possibility of there being other weapons concealed on his person would increase.” 153 U. S. App. D. C., at 150, 471 F. 2d, at 1118 (emphasis in original). But as Chief Judge Bazelon indicated in his opinion below, this kind of reasoning would render meaningless scope limitations on searches. Were one to accept this logic, for example, it would have been reasonable for the police to search the entire house in Chimel v. California, 395 U. S. 752 (1969), for if a weapon had been found somewhere in the house, the arresting officer would have been alerted to search Chimel himself more thoroughly, as the possibility of there being other weapons concealed on his person would arguably have increased.
Nor would it necessarily have been reasonable for the police to have opened the cigarette package at the police station. The Government argued below, as an alternative theory to justify the search in this case, that when a suspect is booked and is about to be placed in station house detention, it is reasonable to search his person to prevent the introduction of weapons or contraband into the jail facility and to inventory the personal effects found on the suspect. Since respondent’s cigarette package would have been removed and opened at the station house anyway, the argument goes, the search might just as well take place in the field at the time of the arrest. This argument fails for two reasons. First, as the Court of Appeals had indicated in its opinion in United States v. Mills, 153 U. S. App. D. C. 156, 472 F. 2d 1231 (1972) (en banc), the justification for station-house searches is not the booking process itself, but rather the fact that the suspect will be placed in jail. In the District of Columbia, petty offenses of the sort involved in the present case are bailable, and, as the Government stipulated in Mills, the normal procedure is for offenders to be advised of the opportunity to post collateral at the station house and to avoid an inventory search unless they are unable or refuse to do so. Id., at 160-161, 472 F. 2d, at 1235-1236. One cannot justify a full search in the field on a subsequent event that quite possibly may never take place.
Second, even had it become necessary to place respondent in confinement, it is still doubtful whether one could justify opening up the cigarette package and examining its contents. The purposes of preventing the introduction of weapons or contraband into the jail facility are fully served simply by removing the package from the prisoner. It is argued that the police must inventory effects found on the prisoner in order to avoid a later claim by the prisoner that jail personnel stole his property. But as the Court of Appeals noted in Mills, the police can protect themselves against such claims by means involving a less extreme intrusion on privacy than would be entailed in opening up and examining the *259contents of all effects found on the person. As an example, the Court of Appeals suggested that the prisoner be given “an opportunity, like that accorded someone given a bathhouse locker for temporary use, to 'check' his belongings in a sealed envelope, perhaps upon executing a waiver releasing the officer of any responsibility.” Id., at 164 n. 11, 472 F. 2d, at 1239 n. 11.
The Government also suggested in oral argument before this Court that it would be administratively inconvenient to require a police officer, after removing an object from an arrestee, to hold on to the object rather than to look inside and determine what it contained. Mere administrative inconvenience, however, cannot justify invasion of Fourth Amendment rights. See Chimel v. California, supra, at 768. One can no doubt imagine cases where the inconvenience might be so substantial as to interfere with the task of transporting the suspect into custody. While these situations might necessitate a different rule, certainly in this case there would have been no inconvenience whatsoever. Officer Jenks could easily have placed the cigarette package in his own pocket or handed it to his partner to hold onto until they reached the precinct station.