delivered the opinion of the Court.

Four months before this Court’s decision in Almeida-Sanchez v. United States, 413 U. S. 266 (1973), respondent was stopped in his automobile by a roving border patrol, and three plastic garbage bags containing 270 pounds of marihuana were found in the trunk of his car by Border Patrol agents. On the basis of this evidence an indictment was returned charging him with a violation of 84 Stat. 1260, 21 U. S. C. § 841 (a) (1). When respondent’s motion to suppress the evidence was denied after a hearing, he stipulated in writing that he “did knowingly and intentionally possess, with intent to distribute, the marijuana concealed in the 1962 Chevrolet which he was driving on February 28, 1973.” 1 The District Court found respondent guilty and imposed sentence. On appeal from that judgment, the Court of Appeals for the Ninth Circuit, sitting en banc, reversed the judgment on the ground that the “rule announced by the Supreme Court in Almeida-Sanchez v. United States . . . should be applied to similar cases pending on appeal on the date the Supreme Court’s decision was announced.” 500 F. 2d 985, 986 (1974) (footnote omitted) .2 We granted the Government’s petition for certiorari. 419 U. S. 993 (1974).

In Almeida-Sanchez, supra, this Court held that a warrantless automobile search, conducted approximately 25 air miles from the Mexican border by Border Patrol agents, acting without probable cause, was uncon*533stitutional under the Fourth Amendment.3 In this case the Government conceded in the Court of Appeals that the search of respondent’s automobile approximately 70 air miles from the Mexican border and the seizure of the marihuana were unconstitutional under the standard announced in Almeida-Sanchez, but it contended that that standard should not be applied to searches conducted prior to June 21, 1973, the date of the decision in AlmeidaSanchez. In an inquiry preliminary to balancing the interests for and against retroactive application, see Stovall v. Denno, 388 U. S. 293, 297 (1967), the majority of the Court of Appeals first considered whether this Court had “articulated a new doctrine” in Almeida-Sanchez, 500 F. 2d, at 987. See, e. g., Chevron Oil Co. v. Huson, 404 U. S. 97, 106 (1971); Milton v. Wainwright, 407 U. S. 371, 381-382, n. 2 (1972) (Stewart, J., dissenting). Concluding that Almeida-Sanchez overruled no prior decision of this Court and instead “reaffirmed well-established Fourth Amendment standards” that did not “disturb a long-accepted and relied-upon practice,” 500 F. 2d, at 988, the Court of Appeals held:

“[Respondent] is entitled to the benefit of the rule announced in Almeida-Sanchez, not because of retro-activity but because of Fourth Amendment principles never deviated from by the Supreme Court.” Id., at 989.

The judgment of conviction was reversed, and the case *534was remanded to the District Court to suppress the evidence seized from respondent’s automobile.

Although expressing some doubt about the applicability of the old law-new law test as a precondition to retro-activity analysis, id., at 990, the six dissenters joined issue with the majority over the proper interpretation of Almeida-Sanchez. The dissenters concluded that AlmeidaSanchez had announced a new constitutional rule because the decision overruled a consistent line of Courts of Appeals precedent and disrupted a long accepted and widely relied upon administrative practice. Border Patrol agents had conducted roving searches pursuant to congressional authorization, 66 Stat. 233, 8 U. S. C. § 1357 (a)(3), and administrative regulation, 8 CFR § 287.1 (a)(2) (1973), which had been continuously upheld until this Court’s decision in Almeida-Sanchez. Since Almeida-Sanchez stated a new rule, the dissenters concluded that the applicability of that decision to pre-June 21, 1973, roving patrol vehicle searches should be determined by reference to the standards summarized in Stovall v. Denno, supra.4 For the reasons expressed in Part II of Judge Wallace’s opinion in United States v. Bowen, 500 F. 2d 960, 975-981 (CA9), cert. granted, 419 U. S. 824 (1974), the dissenters concluded that Almeida-Sanchez should be accorded prospective application.

Despite the conceded illegality of the search under the Almeida-Sanchez standard, the Government contends that the exclusionary rule should not be mechanically applied in the case now before us because the policies *535underlying the rule do not justify its retroactive application to pre-Almeida-Sanchez searches. We agree.

I

Since 1965 this Court has repeatedly struggled with the question of whether rulings in criminal cases should be given retroactive effect. In those cases “[w]here the major purpose of new constitutional doctrine is to overcome an aspect of the criminal trial that substantially impairs its truth-finding function and so raises serious questions about the accuracy of guilty verdicts in past trials,” Williams v. United States, 401 U. S. 646, 653 (1971), the doctrine has quite often been applied retroactively. It is indisputable, however, that in every case in which the Court has addressed the retroactivity problem in the context of the exclusionary rule, whereby coneededly relevant evidence is excluded in order to enforce a constitutional guarantee that does not relate to the integrity of the factfinding process, the Court has concluded that any such new constitutional principle would be accorded only prospective application.5 Linkletter v. Walker, 381 U. S. 618 (1965); Johnson v. New Jersey, 384 U. S. 719 (1966); Stovall v. Denno, supra; Fuller v. Alaska, 393 U. S. 80 (1968); Desist v. United States, 394 U. S. 244 (1969); Jenkins v. Delaware, 395 U. S. 213 *536(1969); Williams v. United States, supra; Hill v. California, 401 U. S. 797 (1971).

We think that these cases tell us a great deal about the nature of the exclusionary rule, as well as something about the nature of retroactivity analysis. Decisions of this Court applying the exclusionary rule to unconstitutionally seized evidence have referred to “the imperative of judicial integrity,” Elkins v. United States, 364 U. S. 206, 222 (1960), although the Court has relied principally upon the deterrent purpose served by the exclusionary rule. See Mapp v. Ohio, 367 U. S. 643 (1961); Lee v. Florida, 392 U. S. 378 (1968); see also United States v. Calandra, 414 U. S. 338 (1974); Michigan v. Tucker, 417 U. S. 433 (1974). And see also Oaks, Studying the Exclusionary Rule in Search and Seizure, 37 U. Chi. L. Rev. 665, 668-672 (1970).

When it came time to consider whether those decisions would be applied retroactively, however, the Court recognized that the introduction of evidence which had been seized by law enforcement officials in good-faith compliance with then-prevailing constitutional norms did not make the courts “accomplices in the willful disobedience of a Constitution they are sworn to uphold.” Elkins v. United States, supra, at 223. Thus, while the “imperative of judicial integrity” played a role in this Court’s decision to overrule Wolf v. Colorado, 338 U. S. 25 (1949), see Mapp v. Ohio, supra, at 659, the Mapp decision was not applied retroactively: “Rather than being abhorrent at the time of seizure in this case, the use in state trials of illegally seized evidence had been specifically authorized by this Court in Wolf.” Linkletter v. Walker, supra, at 638 (footnote omitted). Similarly, in Lee v. Florida, supra, this Court overruled Schwartz v. Texas, 344 U. S. 199 (1952), and held that evidence seized in violation of § 605 of the Federal Communica*537tions Act of 1934, 48 Stat. 1103, 47 IT. S. C. § 605, by state officers could not be introduced into evidence at state criminal trials:

“[T]he decision we reach today is not based upon language and doctrinal symmetry alone. It is buttressed as well by the ‘imperative of judicial integrity.’ Elkins v. United States, 364 U. S. 206, 222. Under our Constitution no court, state or federal, may serve as an accomplice in the willful transgression of ‘the Laws of the United States,’ laws by which ‘the Judges in every State [are] bound ....’” 392 U. S., at 385-386 (footnotes omitted).

But when it came time to consider the retroactivity of Lee, the Court held that it would not be applied retroactively, saying:

“Retroactive application of Lee would overturn every state conviction obtained in good-faith reliance on Schwarts. Since this result is not required by the principle upon which Lee was decided, or necessary to accomplish its purpose, we hold that the exclusionary rule is to be applied only to trials in which the evidence is sought to be introduced after the date of our decision in Lee.” Fuller v. Alaska, supra, at 81.

The teaching of these retroactivity cases is that if the law enforcement officers reasonably believed in good faith that evidence they had seized was admissible at trial, the “imperative of judicial integrity” is not offended by the introduction into evidence of that material even if decisions subsequent to the search or seizure have broadened the exclusionary rule to encompass evidence seized in that manner. It would seem to follow a fortiori from the Linkletter and Fuller holdings that the “im*538perative of judicial integrity” is also not offended if law enforcement officials reasonably believed in good faith that their conduct was in accordance with the law even if decisions subsequent to the search or seizure have held that conduct of the type engaged in by the law enforcement officials is not permitted by the Constitution. For, although the police in Linkletter and Fuller could not have been expected to foresee the application of the exclusionary rule to state criminal trials, they could reasonably have entertained no similar doubts as to the illegality of their conduct. See Wolf v. Colorado, 338 U. S., at 27; § 605 of the Federal Communications Act of 1934; cf. Nardone v. United States, 302 U. S. 379 (1937).

This approach to the “imperative of judicial integrity” does not differ markedly from the analysis the Court has utilized in determining whether the deterrence rationale undergirding the exclusionary rule would be furthered by retroactive application of new constitutional doctrines. See Linkletter v. Walker, supra, at 636-637; Fuller v. Alaska, supra, at 81; Desist v. United States, supra, at 249-251. In Desist, the Court explicitly recognized the interrelation between retroactivity rulings and the exclusionary rule: “[W]e simply decline to extend the court-made exclusionary rule to cases in which its deterrent purpose would not be served.” 394 U. S., at 254 n. 24.

This focus in the retroactivity cases on the purposes served by the exclusionary rule is also quite in harmony with the approach taken generally to the exclusionary rule. In United States v. Calandra, 414 U. S., at 348, we said that the exclusionary rule “is a judicially created remedy designed to safeguard Fourth Amendment rights generally through its deterrent effect, rather than a personal constitutional right of the party aggrieved.” It follows that “the application of the rule has been re*539stricted to those areas where its remedial objectives are thought most efficaciously served.” Ibid. We likewise observed in Michigan v. Tucker, 417 U. S., at 447:

“The deterrent purpose of the exclusionary rule necessarily assumes that the police have engaged in willful, or at the very least negligent, conduct which has deprived the defendant of some right. By refusing to admit evidence gained as a result of such conduct, the courts hope to instill in those particular investigating officers, or in their future counterparts, a greater degree of care toward the rights of an accused. Where the official action was pursued in complete good faith, however, the deterrence rationale loses much of its force.”

The “reliability and relevancy,” Linkletter, supra, at 639, of the evidence found in the trunk of respondent’s car is unquestioned. It was sufficiently damning on the issue of respondent’s guilt or innocence that he stipulated in writing that in effect he had committed the offense charged. Whether or not the exclusionary rule should be applied to the roving Border Patrol search conducted in this case, then, depends on whether considerations of either judicial integrity or deterrence of Fourth Amendment violations are sufficiently weighty to require that the evidence obtained by the Border Patrol in this case be excluded.

II

The Border Patrol agents who stopped and searched respondent’s automobile were acting pursuant to § 287 (a) (3) of the Immigration and Nationality Act of 1952, 66 Stat. 233, 8 U. S. C. § 1357 (a)(3)6 That provision, *540which carried forward statutory authorization dating back to 1946, 60 Stat. 865, 8 U. S. C. § 110 (1946 ed.),7 authorizes appropriately designated Immigration and Naturalization officers to search vehicles “within a reasonable distance from any external boundary of the United States” without a warrant. Pursuant to this statutory authorization, regulations were promulgated fixing the “reasonable distance,” as specified in § 287 (a) (3), at “100 air miles from any external boundary of the United States,” 22 Fed. Reg. 9808 (1957), as amended, 29 Fed. Reg. 13244 (1964), 8 CFR § 287.1 (a) (2) (1973).

Between 1952 and Almeida-Sanchez, roving Border Patrol searches under § 287 (a) (3) were upheld repeatedly against constitutional attack.8 Dicta in many *541other Fifth,9 Ninth,10 and Tenth Circuit11 decisions strongly suggested that the statute and the Border Patrol policy were acceptable means for policing the immigration laws. As Mr. Justice Powell observed in his concurring opinion in Almeida-Sanchez:

“Roving automobile searches in border regions for aliens . . . have been consistently approved by the judiciary. While the question is one of first impression in this Court, such searches uniformly have been sustained by the courts of appeals whose jurisdictions include those areas of the border between Mexico and the United States where the problem has been most severe.” 413 U. S., at 278.

It was in reliance upon a validly enacted statute, supported by longstanding administrative regulations and continuous judicial approval, that Border Patrol agents stopped and searched respondent’s automobile. Since the parties acknowledge that Almeida-Sanchez was the first roving Border Patrol case to be decided by this *542Court, unless we are to hold that parties may not reasonably rely upon any legal pronouncement emanating from sources other than this Court, we cannot regard as blameworthy those parties who conform their conduct to the prevailing statutory or constitutional norm.12 Cf. Chevron Oil Co. v. Huson, 404 U. S. 97 (1971); Lemon v. Kurtzman, 411 U. S. 192 (1973). If the purpose of the exclusionary rule is to deter unlawful police conduct then evidence obtained from a search should be suppressed only if it can be said that the law enforcement officer had knowledge, or may properly be charged with knowledge, that the search was unconstitutional under the Fourth Amendment. Admittedly this uniform treatment of roving border patrol searches by the federal judiciary was overturned by this Court’s decision in AlmeidaSanchez. But in light of this history and of what we perceive to be the purpose of the exclusionary rule, we conclude that nothing in the Fourth Amendment, or in the exclusionary rule fashioned to implement it, requires that the evidence here be suppressed, even if we assume that respondent’s Fourth Amendment rights were violated by the search of his car.13

The judgment of the Court of Appeals is therefore

Reversed.

dissenting.

I agree with my Brother Brennan that Almeida-Sanchez v. United States, 413 U. S. 266 (1973), reaffirmed traditional Fourth Amendment principles and that the purposes of the exclusionary rule compel exclusion of the unconstitutionally seized evidence in this case. I adhere to my view that a constitutional rule made retroactive in one case must be applied retroactively in all. See my dissent in Daniel v. Louisiana, 420 U. S. 31, 33 (1975), and cases cited. It is largely a matter of chance that we held the Border Patrol to the command of the Fourth Amendment in Almeida-Sanchez rather than in the case of this defendant. Equal justice does not permit a defendant’s fate to depend upon such a fortuity. The judgment below should be affirmed.

with, whom Mr. Justice Marshall joins, dissenting.

I

Until today the question of the prospective application of a decision of this Court was not deemed to be presented unless the decision “constitute[d] a sharp break in the line of earlier authority or an avulsive change which caused the current of the law thereafter to flow between new banks.” Hanover Shoe, Inc. v. United Shoe Machinery Corp., 392 U. S. 481, 499 (1968).1 Measured by that test, our decision in Almeida-Sanchez v. United States, 413 U. S. 266 (1973), presents no question of prospectivity, and the Court errs in even addressing the question. For both the Court’s opinion and the concurring opinion of Mr. Justice Powell in AlmeidaSanchez plainly applied familiar principles of constitutional adjudication announced 50 years ago in Carroll v. United States, 267 U. S. 132, 153-154 (1925), and merely construed 66 Stat. 233, 8 U. S. C. § 1357 (a)(3), so as to render it constitutionally consistent with that decision. 413 U. S., at 272; id., at 275, and n. 1 (Powell, J., concurring).

The Court states, however, that the Border Patrol agents searched Peltier “in reliance upon a validly enacted statute, supported by longstanding administrative regulations and continuous judicial approval. . . .” Ante, *545at 541. With all respect, any such reliance would be misplaced. First, the Court repeats the error of my Brother White in his dissent in Almeida-Sanchez in finding express congressional and administrative approval for random roving patrol searches. 413 U. S., at 291, 292-293, 296. The statute, 8 U. S. C. § 1357 (a), only authorizes searches of vehicles “without warrant . . . within a reasonable distance from any external boundary” ; nothing in the statute expressly dispenses with the necessity for showing probable cause. The regulation, 8 CFR § 287.1 (a) (2) (1973), merely defined “a reasonable distance” as “within 100 air miles”; it, too, does not purport to exempt the Border Patrol from observing the probable-cause requirement.2

Second, the Court states that “[b]etween 1952 and Almeida-Sanchez, roving Border Patrol searches under § 287 (a)(3) were upheld repeatedly against constitutional attack.” Ante, at 540. But the first decision of the Court of Appeals for the Ninth Circuit squarely in point, United States v. Miranda, 426 F. 2d 283, was decided in 1970, and the second, United States v. Almeida*546Sanchez, 452 F. 2d 459, was decided over strong dissent in 1971 and was pending on certiorari in this Court when Peltier was searched. 406 U. S. 944 (1972). The first decision of the Court of Appeals for the Tenth Circuit approving alien searches by roving patrols without either probable cause or any suspicious conduct was in 1969. Roa-Rodriquez v. United States, 410 F. 2d 1206. And the Court of Appeals for the Fifth Circuit, unlike the Ninth and Tenth Circuits, always required at least a “reasonable suspicion” that a car might contain aliens as the basis of a valid search under 8 U. S. C. § 1357 (a) (3). United States v. Wright, 476 F. 2d 1027, 1030, and n. 2 (1973), and cases cited.

In addition, the rule of Miranda, supra, was a patent anomaly in the Courts of Appeals which sanctioned roving patrol searches without a showing even of suspicious circumstances. The Court of Appeals for the Ninth Circuit, for example, held consistently that probable cause must be shown to validate a search for contraband except in a border search or its functional equivalent, see, e. g., Cervantes v. United States, 263 F. 2d 800, 803 (1959); Fumagalli v. United States, 429 F. 2d 1011 (1970),3 and this despite a statutory authorization to search for contraband at least as broad as § 1357 (a) *547(3). See 14 Stat. 178, 19 U. S. C. § 482.4 Moreover, the Courts of Appeals require some measure of cause to suspect violation of law in interrogations and arrests authorized by other subsections of 8 U. S. C. § 1357 (a). See Au Yi Lau v. INS, 144 U. S. App. D. C. 147, 445 F. 2d 217 (1971); Yam Sang Kwai v. INS, 133 U. S. App. D. C. 369, 411 F. 2d 683 (1969).

Given this history, it becomes quite clear why the Court has found it necessary to discard the “sharp break” test to reach the prospectivity question in this case. For the approval by Courts of Appeals of this law enforcement practice was short-lived, less than unanimous, irreconcilable with other rulings of the same courts, and contrary to the explicit doctrine of this Court in Carroll, supra, as reaffirmed in Brinegar v. United States, 338 U. S. 160, 164 (1949), and other cases. If a case in this Court merely reaffirming longstanding precedent can ever constitute the “avulsive change [in] the current of the law” required before we even address the issue of prospectivity, Hanover Shoe, 392 U. S., at 499, surely Almeida-Sanchez was not such a case.5

*548This case is a good illustration of the dangers of addressing prospectivity where the “sharp break” standard is not met. As this Court has recognized, applying a decision only prospectively,6 can entail inequity to others whose cases are here on direct review but are held pending decision of the case selected for decision. Stovall v. Denno, 388 U. S. 293, 301 (1967). Although I continue to believe that denial of the benefits of the decision in such cases is a tolerable anomaly in cases in which defendants *549were accorded all constitutional rights then announced by this Court, it becomes intolerable, and a travesty of justice, when the Court does no more than reaffirm and apply long-established constitutional principles to correct an aberration created by the courts of appeals.

More fundamentally, applying a decision of this Court prospectively when the decision is not a “sharp break in the web of the law,” Milton v. Wainwright, 407 U. S. 371, 381 n. 2 (1972) (Stewart, J., dissenting), encourages in those responsible for law enforcement a parsimonious approach to enforcement of constitutional rights. “One need not be a rigid partisan of Blackstone to recognize that many, though not all, of this Court’s constitutional decisions are grounded upon fundamental principles whose content does not change dramatically from year to year . . . .” Desist v. United States, 394 U. S. 244, 263 (1969) (Harlan, J., dissenting). To apply our opinions prospectively except in “sharp break” cases “add[s] this Court’s approval to those who honor the Constitution’s mandate only where acceptable to them or compelled by the precise and inescapable specifics of a decision of this Court. . . . History does not embrace the years needed for us to hold, millimeter by millimeter, that such and such a penetration of individual rights is an infringement of the Constitution’s guarantees. The vitality of our Constitution depends upon conceptual faithfulness and not merely decisional obedience. Certainly, this Court should not encourage police or other courts to disregard the plain purport of our decisions and to adopt a let’s-wait-until-it’s-decided approach.” Id., at 277 (Fortas, J., dissenting).7

*550II

Nevertheless, the Court substitutes, at least as respects the availability of the exclusionary rule in cases involving searches invalid under the Fourth Amendment, a presumption against the availability of decisions of this Court except prospectively. The substitution discards not only the “sharp break” determinant but also the equally established principle that prospectivity “is not automatically determined by the provision of the Constitution on which the dictate is based. . . . [W]e must determine retroactivity ‘in each case’ by looking to the peculiar traits of the specific ‘rule in question.’ ” Johnson v. New Jersey, 384 U. S. 719, 728 (1966).8 Link-letter v. Walker, 381 U. S. 618 (1965), the seminal prospectivity decision, held only that “the Court may in the interest of justice make [a] rule prospective . . . where the exigencies of the situation require such an application.” Id., at 628 (emphasis added). Today the Court stands the Linkletter holding on its head by creating a class of cases in which nonretroactivity is the rule and not, as heretofore, the exception.

The Court’s stated reason for this remarkable departure from settled principles is “the policies underlying the [exclusionary] rule.” Ante, at 534-535. But the policies identified by the Court as underlying that rule in Fourth Amendment cases are distorted out of all resemblance to the understanding of purposes that has heretofore prevailed. I said in my dissent in United States v. Calandra, 414 U. S. 338 (1974), that that decision left' *551me “with the uneasy feeling that ... a majority of my colleagues have positioned themselves to . . . abandon altogether the exclusionary rule in search-and-seizure cases.” Id., at 365. My uneasiness approaches conviction after today’s treatment of the rule.

Ill

The Court’s opinion depends upon an entirely new understanding of the exclusionary rule in Fourth Amendment cases, one which, if the vague contours outlined today are filled in as I fear they will be, forecasts the complete demise of the exclusionary rule as fashioned by this Court in over 61 years of Fourth Amendment jurisprudence. See Weeks v. United States, 232 U. S. 383 (1914).9 An analysis of the Court’s unsuccessfully veiled reformulation demonstrates that its apparent rush to discard 61 years of constitutional development has produced a formula difficult to comprehend and, on any understanding of its meaning, impossible to justify.

The Court signals its new approach in these words: “If the purpose of the exclusionary rule is to deter unlawful police conduct, then evidence obtained from a search should be suppressed only if it can be said that the law enforcement officer had knowledge, or may properly be charged with knowledge, that the search was unconstitutional under the Fourth Amendment.” Ante, at 542. True, the Court does not state in so many words that this formulation of the exclusionary rule is to be applied beyond the present retroactivity context. But the proposition is stated generally and, particularly in view of *552the concomitant expansion of prospectivity announced today, Part I, supra, I have no confidence that the new formulation is to be confined to putative retroactivity cases. Rather, I suspect that when a suitable opportunity arises, today’s revision of the exclusionary rule will be pronounced applicable to all search-and-seizure cases. I therefore register my strong disagreement now.

The new formulation obviously removes the very foundation of the exclusionary rule as it has been expressed in countless decisions. Until now the rule in federal criminal cases decided on direct review10 has been that suppression is necessarily the sanction to be applied when it is determined that the evidence was in fact illegally acquired.11 The revision unveiled today *553suggests that instead of that single inquiry, district judges may also have to probe the subjective knowledge of the official who orders the search, and the inferences from existing law that official should have drawn.12 The decision whether or not to order suppression would then turn upon whether, based on that expanded inquiry, suppression would comport with either the deterrence rationale of the exclusionary rule or “the imperative of judicial integrity.” 13

*554On this reasoning, Almeida-Sanchez itself was wrongly-decided. For if the Border Patrolmen who searched Peltier could not have known that they were acting unconstitutionally, and thus could not have been deterred from the search by the possibility of the exclusion of the evidence from the trial, obviously the Border Patrolmen who searched Almeida-Sanchez several years earlier had no reason to be any more percipient. If application of the exclusionary rule depends upon a showing that the particular officials who conducted or authorized a particular search knew or should have known that they were violating a specific, established constitutional right, the reversal of Almeida-Sanchez’ conviction was plainly error.

Other defects of today’s new formulation are also patent. First, this new doctrine could stop dead in its tracks judicial development of Fourth Amendment rights. For if evidence is to be admitted in criminal trials in the absence of clear precedent declaring the search in question unconstitutional, the first duty of a court will be to deny the accused’s motion to suppress if he cannot cite a case invalidating a search or seizure on identical facts.14 Yet, even its opponents concede *555that the great service of the exclusionary rule has been its usefulness in forcing judges to enlighten our understanding of Fourth Amendment guarantees. “It is . . . imperative to have a practical procedure by which courts can review alleged violations of constitutional rights and articulate the meaning of those rights. The advantage of the exclusionary rule — entirely apart from any direct deterrent effect — is that it provides an occasion for judicial review, and it gives credibility to the constitutional guarantees. By demonstrating that society will attach serious consequences to the violation of constitutional rights, the exclusionary rule invokes and magnifies the moral and educative force of the law. Over the long term this may integrate some fourth amendment ideals into the value system or norms of behavior of law enforcement agencies.” Oaks, Studying the Exclusionary Rule in Search and Seizure, 37 U. Chi. L. Rev. 665, 756 (1970) (hereafter Oaks). See also Amsterdam, Perspectives on the Fourth Amendment, 58 Minn. L. Rev. 349, 429-430 (1974) (hereafter Amsterdam) . While distinguished authority has suggested that an effective affirmative remedy could equally serve that function, see Oaks, supra, and Bivens v. Six Unknown Federal Narcotics Agents, 403 U. S. 388, 420-423 (1971) (Burger, C. J., dissenting), no equally effective alternative has yet been devised.

*556Second, contrary to the Court’s assumption, the exclusionary rule does not depend in its deterrence rationale on the punishment of individual law enforcement officials.15 Indeed, one general fallacy in the reasoning of critics of the exclusionary rule is the belief that the rule is meant to deter official wrongdoers by punishment or threat of punishment. It is also the fallacy of the Court’s attempt today to outline a revision in the exclusionary rule.

Deterrence can operate in several ways. The simplest is special or specific deterrence — -punishing an individual so that he will not repeat the same behavior. But “[t]he exclusionary rule is not aimed at special deterrence since it does not impose any direct punishment on a law enforcement official who has broken the rule. . . . The exclusionary rule is aimed at affecting the wider audience of all law enforcement officials and society at *557large. It is meant to discourage violations by individuals who have never experienced any sanction for them.” Oaks 709-710.16

Thus, the exclusionary rule, focused upon general, not specific, deterrence, depends not upon threatening a sanction for lack of compliance but upon removing an inducement to violate Fourth Amendment rights. Elkins v. United States, 364 U. S. 206, 217 (1960), clearly explained that the exclusionary rule’s “purpose is to deter— to compel respect for the constitutional guaranty in the only effectively available way — by removing the incentive to disregard it.” (Emphasis added.) “A criminal court system functioning without an exclusionary rule ... is the equivalent of a government purchasing agent paying premium prices for evidence branded with the stamp of unconstitutionality. ... If [the Government] receives the products of [illegal] searches and seizures . . . and uses them as the means of convicting people whom the officer conceives it to be his job to get convicted, it is not merely tolerating but inducing unconstitutional *558searches and seizures.” Amsterdam 431-432.17 (Emphasis supplied.)

We therefore might consider, in this light, what may have influenced the officials who authorized roving searches without probable cause under the supposed authority of 8 U. S. C. § 1357 (a) (3) and 8 CFR § 287.1 (a) (2) (1973).18 The statute is at best ambiguous as to *559whether probable cause is required, though quite explicit that a warrant is not.19 The officials could therefore read the statute in one of two ways. They could read it not to require probable cause, regard as irrelevant Carroll v. United States, 267 U. S. 132 (1925), requiring probable cause, though no warrant, before stopping and searching a moving automobile unless the search is at the border, and command their subordinates to stop at random any car within 100 miles of the border and search for illegal aliens. Or they could conclude that because the statute is silent about probable cause, and because Carroll seems to require it, they should instruct their subordinates to stop moving vehicles away from the border only if there is some good reason to believe that they contain illegal aliens. Obviously, today’s decision is a wide-open invitation to pursue the former course, because if this Court later decides that the officers guessed wrong in a particular case, one conviction will perhaps be lost, but many will have been gained, see supra, at 549, 554. The concept of the exclusionary rule until today, however, was designed to discourage officials from invariably opting for the choice that compromises Fourth Amendment rights, even though that rule has not worked perfectly as it did not in this case. “The efforts of the courts and their officials to bring the guilty to punishment, praiseworthy as they are, are not to be aided by the sacrifice of those great principles established by years of endeavor and *560suffering which have resulted in their embodiment in the fundamental law of the land.” Weeks, 232 U. S., at 393 (emphasis supplied).

Aside from this most fundamental error, solid practical reasons militate forcefully in favor of rejection of today’s suggested road to revision of the exclusionary rule. This Court has already rejected a case-by-case approach to the exclusionary rule. After Wolf v. Colorado, 338 U. S. 25 (1949), had held the Fourth Amendment applicable to the States without also requiring the States to follow the exclusionary rule of Weeks, Irvine v. California, 347 U. S. 128 (1954), presented the opportunity of compelling the States to apply Weeks in especially egregious situations such as Irvine’s. The Court rejected the opportunity because “a distinction of the kind urged would leave the rule so indefinite that no state court could know what it should rule in order to keep its processes on solid constitutional ground.” Id., at 134 (opinion of Jackson, J.). See also id., at 138 (Clark, J., concurring).

Today’s formulation extended to all search-and-seizure cases would inevitably introduce the same uncertainty, by adding a new layer of factfinding in deciding motions to suppress in the already heavily burdened federal courts. The district courts would have to determine, and the appellate courts to review, subjective states of mind of numerous people, see n. 18, supra, and reasonable objective extrapolations of existing law, on each of the thousands of suppression motions presented each year.20 Nice questions will have to be faced, such as whether to exclude evidence obtained in a search which officers be*561lieved to be unconstitutional but which in fact was not, and whether to exclude evidence obtained in a search in fact unconstitutional and believed to be unconstitutional, but which the ordinary, reasonable police officer might well have believed was constitutional. One criticism of the present formulation of the exclusionary rule is that it may deflect the inquiry in a criminal trial from the guilt of the defendant to the culpability of the police. The formulation suggested today would vastly exacerbate this possibility, heavily burden the lower courts, and worst of all, erode irretrievably the efficacy of the exclusion principle.21 Indeed, “no [federal] court could know what it should rule in order to keep its processes on solid constitutional ground.” Cf. 347 U. S., at 134. Because of the superficial and summary way that the Court treats the question the formulation will, I am certain, be unsatisfactory even to those convinced, as I am not, that the exclusionary rule must be drastically overhauled.22

If a majority of my colleagues are determined to discard the exclusionary rule in Fourth Amendment cases, they should forthrightly do so, and be done with it. This business of slow strangulation of the rule, with no *562opportunity afforded parties most concerned to be heard, would be indefensible in any circumstances. But to attempt covertly the erosion of an important principle over 61 years in the making as applied in federal courts clearly demeans the adjudicatory function, and the institutional integrity of this Court.