concurring in the judgments in Nos. 36 and 82 and dissenting in No. 81.
These three cases have one question in common: the extent to which new constitutional rules prescribed by this Court for the conduct of criminal cases are applicable to other such cases which were litigated under different but then-prevailing constitutional rules.
One of these cases is before us on direct review, No. 81, Williams, the other two being here on collateral review, No. 82, Elkanich, and No. 36, Mackey. In each instance the new rule is held not applicable, and, in *676consequence, the judgments below are affirmed, without reaching the merits of the underlying questions presented. Two of the cases, Williams and Elkanich, involve the Court’s decision in Chimel v. California, 395 U. S. 752 (1969), changing the rule as to the scope of permissible searches and seizures incident to a lawful arrest. The other case, Mackey, involves the Court’s decisions in Marchetti v. United States, 390 U. S. 39 (1968), and Grosso v. United States, 390 U. S. 62 (1968), changing the rule as to the application of the privilege against self-incrimination with respect to criminal prosecutions arising under the federal gambling tax statutes.
Today’s decisions mark another milestone in the development of the Court’s “retroactivity” doctrine, which came into being somewhat less than six years ago in Linkletter v. Walker, 381 U. S. 618 (1965). That doctrine was the product of the Court’s disquietude with the impacts of its fast-moving pace of constitutional innovation in the criminal field. Some members of the Court, and I have come to regret that I was among them, initially grasped this doctrine as a way of limiting the reach of decisions that seemed to them fundamentally unsound. Others rationalized this resort to prospectivity as a “technique” that provided an “impetus ... for the implementation of long overdue reforms, which otherwise could not be practicably effected.” Jenkins v. Delaware, 395 U. S. 213, 218 (1969). The upshot of this confluence of viewpoints was that the subsequent course of Linkletter became almost as difficult to follow as the tracks made by a beast of prey in search of its intended victim. See my dissenting opinion in Desist v. United States, 394 U. S. 244, 256-257 (1969). See also United States v. United States Coin & Currency, post, p. 728 (appendix to concurring opinion of Brennan, J.). It was this train of events that impelled me to suggest two Terms ago in Desist that the time had come for us *677to pause to consider just where these haphazard developments might be leading us. That is. what I had thought underlay the taking of these cases, and their companions, United States v. United States Coin & Currency, post, p. 715, and Hill v. California, post, p. 797. Regrettably, however, this opportunity has largely eventuated in little more than a reaffirmation of the post-Linkletter developments.
What emerges from today’s decisions is that in the realm of constitutional adjudication in the criminal field the Court is free to act, in effect, like a legislature, making its new constitutional rules wholly or partially retroactive or only prospective as it deems wise. I completely disagree with this point of view. While I do not subscribe to the Blackstonian theory that the law should be taken to have always been what it is said to mean at a later time, I do believe that whether a new constitutional rule is to be given retroactive or simply prospective effect must be determined upon principles that comport with the judicial function, and not upon considerations that are appropriate enough for a legislative body.
I
At the outset, I think it is clear that choosing a binding, generally applicable interpretation of the Constitution presents a problem wholly different from that of choosing whether to apply the rule so evolved “retroactively” to other cases arising on direct review.
In adopting a particular constitutional principle, this Court very properly weighs the nature and purposes of various competing alternatives, including the extent to which a proposed rule will enhance the integrity of the criminal process and promote the efficient administration of justice, as well as the extent to which justifiable expectations have grown up surrounding one rule or another. Indeed, it is this very process of weighing such *678factors that should constitute the core of our task in giving concrete embodiment to those constitutional commands that govern the procedures by which the State and Federal Governments enforce their criminal laws.
But we possess this awesome power of judicial review, this duty to bind coordinate branches of the federal system with our view of what the Constitution dictates, only because we are a court of law, an appellate court charged with the responsibility of adjudicating cases or controversies according to the law of the land and because the law applicable to any such dispute necessarily includes the Federal Constitution. That is the classic explanation for the basis of judicial review, an explanation first put forth by Chief Justice Marshall in Marbury v. Madison, 1 Cranch 137, 177-178 (1803), and from that day to this the sole continuing rationale for the exercise of this judicial power:
“Certainly all those who have framed written constitutions contemplate them as forming the fundamental and paramount law of the nation ....
“It is emphatically the province and duty of the judicial department to say what the law is. Those who apply the rule to particular cases, must of necessity expound and interpret that rule. If two laws conflict with each other, the courts must decide on the operation of each.
“If then the courts are to regard the constitution; and the constitution is superior to any ordinary act of the legislature; the constitution, and not such ordinary act, must govern the case to which they both apply.”
We announce new constitutional rules, then, only as a correlative of our dual duty to decide those cases over *679which we have jurisdiction and to apply the Federal Constitution as one source of the matrix of governing legal rules. We cannot release criminals from jail merely because we think one case is a particularly appropriate one in which to apply what reads like a general rule of law or in order to avoid making new legal norms through promulgation of dicta. This serious interference with the corrective process is justified only by necessity, as part of our task of applying the Constitution to cases before us. Simply fishing one case from the stream of appellate review, using it as a vehicle for pronouncing new constitutional standards, and then permitting a stream of similar cases subsequently to flow by unaffected by that new rule constitute an indefensible departure from this model of judicial review.
If we do not resolve all cases before us on direct review in light of our best understanding of governing constitutional principles, it is difficult to see why we should so adjudicate any case at all. If there is no need for an anti-majoritarian judicial control over the content of our legal system in nine cases precisely like that presented by Mr. Chimel’s dispute with the State of California, it is hard to see the necessity, wisdom, or justification for imposing that control in the Chimel case itself. In truth, the Court’s assertion of power to disregard current law in adjudicating cases before us that have not already run the full course of appellate review, is quite simply an assertion that our constitutional function is not one of adjudication but in effect of legislation. We apply and definitively interpret the Constitution, under this view of our role, not because we are bound to, but only because we occasionally deem it appropriate, useful, or wise. That sort of choice may permissibly be made by a legislature or a council of revision, but not by a court of law.
The notion that cases before us on direct review need not be adjudicated in accordance with those legal princi-*680pies governing at the time we are possessed of jurisdiction in the case entails additional significant untoward consequences. By this doctrine all other courts in this country are, in effect, reduced largely to the role of automatons, directed by us to apply mechanistically all then-settled federal constitutional concepts to every case before them. No longer do the inferior courts — and, in the constitutional realm, all courts are inferior to us — bear responsibility for developing or interpreting the Constitution. For it is a necessary corollary of this current retroactivity doctrine that an inferior court errs when it arrives at a result which this Court subsequently adopts but later decides must operate prospectively only. See my dissent in Desist, 394 U. S., at 259. Cf. United States v. White, post, p. 754 (Part II), and my dissenting opinion in that case, post, p. 768. See also United States v. United States Coin & Currency, post, p. 730 (White, J., dissenting). Although it is necessary for the proper functioning of the federal system that this Court possess the last word on issues of federal constitutional law, it is intolerable that we take to ourselves the sole ability to speak to such problems.
Refusal to apply new constitutional rules to all cases arising on direct review may well substantially deter those whose financial resources are barely sufficient to withstand the costs of litigating to this Court, or attorneys who are willing to make sacrifices to perform their professional obligation in its broadest sense, from asserting rights bottomed on constitutional interpretations different from those currently prevailing in this Court. More importantly, it tends to cut this Court loose from the force of precedent, allowing us to restructure artificially those expectations legitimately created by extant law and thereby mitigate the practical force of stare decisis, Linkletter v. Walker, 381 U. S., at 644 (Black, J., dissenting), a force which ought properly to bear on the *681judicial resolution of any legal problem. Cf. Moragne v. States Marine Lines, 398 U. S. 375, 403 (1970).
One could catalogue virtually ad infinitum what I view as unacceptable ancillary consequences of this aspect of the Court’s ambulatory retroactivity doctrine. For me, the fact that this doctrine entails an inexplicable and unjustifiable departure from the basic principle upon which rests the institution of judicial review is sufficient to render it untenable. I continue to believe that a proper perception of our duties as a court of law, charged with applying the Constitution to resolve every legal dispute within our jurisdiction on direct review, mandates that we apply the law as it is at the time, not as it once was. Inquiry into the nature, purposes, and scope of a particular constitutional rule is essential to the task of deciding whether that rule should be made the law of the land. That inquiry is, however, quite simply irrelevant in deciding, once a rule has been adopted as part of our legal fabric, which cases then pending in this Court should be governed by it.
II
Of the cases presently under discussion, only Williams involves direct review of a nonfinal criminal judgment. The other two, Elkanich and Mackey, were brought here by persons in federal custody, seeking release through issuance of a writ of habeas corpus.1 At the time their *682convictions became final, there was no constitutional error in the conviction of either. Since that time subsequent decisions of this Court have formulated new constitutional rules that invalidate the procedures like those involved in their trials.
While, as I have just stated, I think it clear what law should be applied to nonfinal convictions here on direct review, the choice of law problem as it applies to cases here on habeas seems to me a much more difficult one. However, that choice, in my view, is also one that can be responsibly made only by focusing, in the first instance, on the nature, function, and scope of the adjudicatory process in which such cases arise. The relevant frame of reference, in other words, is not the purpose of the new rule whose benefit the petitioner seeks, but instead the purposes for which the writ of habeas corpus is made available.
As I first pointed out in my dissent in Desist, 394 U. S., at 260-261, this Court's function in reviewing a decision allowing or disallowing a writ of habeas corpus is, and always has been, significantly different from our role in reviewing on direct appeal the validity of non-final criminal convictions. While the entire theoretical underpinnings of judicial review and constitutional supremacy dictate that federal courts having jurisdiction on direct review adjudicate every issue of law, including federal constitutional issues, fairly implicated by the trial process below and properly presented on appeal, federal courts have never had a similar obligation on habeas corpus.
Habeas corpus always has been a collateral remedy, providing an avenue for upsetting judgments that *683have become otherwise final. It is not designed as a substitute for direct review. The interest in leaving concluded litigation in a state of repose, that is, reducing the controversy to a final judgment not subject to further judicial revision, may quite legitimately be found by those responsible for defining the scope of the writ to outweigh in some, many, or most instances the competing interest in readjudicating convictions according to all legal standards in effect when a habeas petition is filed. Indeed, this interest in finality might well lead to a decision to exclude completely certain legal issues, whether or not properly determined under the law prevailing at the time of trial, from the cognizance of courts administering this collateral remedy. This has always been the case with collateral attacks on final civil judgments.2 More immediately relevant here is the fact that *684prior to Brown v. Allen, 344 U. S. 443 (1953), federal courts would never consider the merits of a constitutional claim raised on habeas if the petitioner had a fair opportunity to raise his arguments in the original criminal proceeding, see my dissent in Fay v. Noia, 372 U. S. 391, 449-463 (1963), unless the petitioner attacked the constitutionality of the federal, Ex parte Siebold, 100 U. S. 371 (1880), or state, Crowley v. Christensen, 137 U. S. 86 (1890), statute under which he had been convicted. See generally Bator, Finality in Criminal Law and Federal Habeas Corpus for State Prisoners, 76 Harv. L. Rev. 441, 463 (1963); Note, Developments in the Law-Federal Habeas Corpus, 83 Harv. L. Rev. 1040, 1042-1062 (1970).
Thus, prior to Brown v. Allen, it must have been crystal clear that the “retroactivity” of a new constitutional rule was a function of the scope and purposes of the habeas corpus writ. Absent unusual circumstances, a new rule was not cognizable on habeas simply because of the limited scope of the writ. While the extent of inquiry into alleged constitutional error on habeas has been drastically expanded in the past 20 years, the retroactivity problem remains analytically constant. In my view, the issues respectively presented by the two cases I treat here that arise on collateral review (Elkanich and Mackey)— whether the new rules of the Chimel case and the Mar-chetti and Grosso cases should be applied “retroactively” — must be considered as none other than a problem as to the scope of the habeas writ. We can properly decline to apply the Chimel rule, or the principles of Marchetti and Grosso, to the present cases only if that is consistent with the reasons for the provision, in our federal legal system, of a habeas corpus proceeding to test the validity of an individual’s official confinement.
Thus I am led to make some inquiry into the purposes of habeas. At the outset I must note that this faces *685me with difficult problems. I have consistently protested a long course of habeas decisions in this Court which, I still believe, constitute an unsound extension of the historic scope of the writ and an unfortunate display of insensitivity to the principles of federalism which underlie the American legal system. See, e. g., Fay v. Noia, 372 U. S. 391, 448 (1963); Sanders v. United States, 373 U. S. 1, 23 (1963); Kaufman v. United States, 394 U. S. 217, 242 (1969); Townsend v. Sain, 372 U. S. 293, 325 (1963) (Stewart, J., dissenting). If I felt free to decide the present cases consistently with my own views of the legitimate role of the Great Writ, I should have little difficulty. But as my views on this score have not commended themselves to most of my Brethren, I feel obliged to approach these two collateral cases within the framework of current habeas corpus doctrine. This is not an easy exercise, for present habeas corpus decisions provide little assistance in fathoming the underlying understanding of habeas corpus upon which these decisions have been premised. The short of the matter is that this Court has in recent times yet to produce any considered, coherent statement of the general purposes of habeas. In considering the problem of “retroactivity” on direct review, it is possible to work from a general classic theory of judicial review, but while the specific uses of the habeas writ have greatly multiplied, the earlier perception of its general metes and bounds has been swallowed up and gone unreplaced. About the only way to proceed is to work from the bottom up, ascertaining first which issues are cognizable on habeas, and which are not, and thereafter inferring what must be thought to be the nature of the writ.
I start with the proposition that habeas lies to inquire into every constitutional defect in any criminal trial, where the petitioner remains “in custody” because of the judgment in that trial, unless the error committed was *686knowingly and deliberately waived or constitutes mere harmless error. That seems to be the implicit premise of Brown v. Allen, supra, and the clear purport of Kaufman v. United States, supra. This is not to say, however, that the function of habeas corpus is to provide a federal forum for determining whether any individual is presently “in custody in violation of the constitution ... of the United States,” 28 U. S. C. §2254 (1964 ed., Supp. V), in the sense that the basis for his incarceration would, under the law existing at the time a petition is filed or adjudicated, as distinguished from the law that was applicable at the time his conviction became final, be held free of constitutional error. Cf. Meador, Habeas Corpus and the “Retroactivity” Illusion, 50 Va. L. Rev. 1115 (1964).
While it has been generally, although not universally, assumed that habeas courts should apply current constitutional law to habeas petitioners before them,3 I do not believe this is or should be the correct view. First, no such proposition has ever been squarely considered and embraced by this Court, at least since the recent proliferation of criminal defendants’ protected constitutional *687rights and the concomitant expansion of the writ.4 Moreover, applying current constitutional standards to convictions finalized while different views were ascendant appears unnecessary to achieve the ends sought by Brown and Kaufman. The primary justification given by the Court for extending the scope of habeas to all alleged constitutional errors is that it provides a quasi-appellate review function, forcing trial and appellate courts in both the federal and state system to toe the constitutional mark. See Kaufman v. United States, 394 U. S., at 226. However, the opinion in Kaufman itself concedes that there is no need to apply new constitutional rules on habeas to serve the interests promoted by that decision. 394 U. S., at 229. Further, as I explain in the margin below,5 Congress, in at least one significant *688regard, seems plainly to have disapproved the notion that supervening constitutional interpretation ought to apply on habeas involving state convictions.
Clearly, it is at least fair to regard this issue as not yet settled by this Court. Consequently, I go on to inquire how it ought to be resolved. For me, with a few exceptions, the relevant competing policies properly balance out to the conclusion that, given the current broad scope of constitutional issues cognizable on habeas, *689it is sounder, in adjudicating habeas petitions, generally to apply the law prevailing at the time a conviction became final than it is to seek to dispose of all these cases on the basis of intervening changes in constitutional interpretation.
I do not mean to neglect the force of countervailing contentions. Assuring every state and federal prisoner a forum in which he can continually litigate the current constitutional validity of the basis for his conviction tends to assure a uniformity of ultimate treatment among prisoners; provides a method of correcting abuses now, but not formerly, perceived as severely detrimental to societal interests; and tends to promote a rough form of justice, albeit belated, in the sense that current constitutional notions, it may be hoped, ring more “correct” or “just” than those they discarded.
In my view, however, these interests are too easily overstated. Some discrimination must always exist in the legal treatment of criminal convicts within a system where the governing law is continuously subject to change. And it has been the law, presumably for at least as long as anyone currently in jail has been incarcerated, that procedures utilized to convict them must have been fundamentally fair, that is, in accordance with the command of the Fourteenth Amendment that “[n]o State shall . . . deprive any person of life, liberty, or property, without due process of law.” Twining v. New Jersey, 211 U. S. 78 (1908). Moreover, it is too easy to suggest that constitutional updating is necessary in order to assure that the system arrives only at “correct” results. By hypothesis, a final conviction, state or federal, has been adjudicated by a court cognizant of the Federal Constitution and duty bound to apply it. To argue that a conclusion reached by one of these “inferior” courts is somehow forever erroneous because years later this Court took a different view of the relevant constitutional com*690mand carries more emotional than analytic force. No one has put this point better than Mr. Justice Jackson, in his concurring opinion in Brown v. Allen, 344 U. S., at 540:
“[RJeversal by a higher court is not proof that justice is thereby better done. There is no doubt that if there were a super-Supreme Court, a substantial proportion of our reversals of state courts would also be reversed. We are not final because we are infallible, but we are infallible only because we are final.”
More importantly, there are operative competing policies in this area which I regard as substantial. It is, I believe, a matter of fundamental import that there be a visible end to the litigable aspect of the criminal process. Finality in the criminal law is an end which must always be kept in plain view. See, e. g., Fay v. Noia, 372 U. S., at 445 (Clark, J., dissenting); Spencer v. Texas, 385 U. S. 554, 583 (1967) (Warren, C. J., concurring and dissenting). See also Bator, Finality in Criminal Law and Federal Habeas Corpus for State Prisoners, 76 Harv. L. Rev. 441 (1963); Friendly, Is Innocence Irrelevant? Collateral Attack on Criminal Judgments, 38 U. Chi. L. Rev. 142, 146-151 (1970). As I have stated before, “Both the individual criminal defendant and society have an interest in insuring that there will at some point be the certainty that comes with an end to litigation, and that attention will ultimately be focused not on whether a conviction was free from error but rather on whether the prisoner can be restored to a useful place in the community.” Sanders v. United States, 373 U. S., at 24-25 (Harlan, J., dissenting). At some point, the criminal process, if it is to function at all, must turn its attention from whether a man ought properly to be incarcerated to how he is to *691be treated once convicted. If law, criminal or otherwise, is worth having and enforcing, it must at some time provide a definitive answer to the questions litigants present or else it never provides an answer at all. Surely it is an unpleasant task to strip a man of his freedom and subject him to institutional restraints. But this does not mean that in so doing, we should always be halting or tentative. No one, not criminal defendants, not the judicial system, not society as a whole is benefited by a judgment providing a man shall tentatively go to jail today, but tomorrow and every day thereafter his continued incarceration shall be subject to fresh litigation on issues already resolved.
A rule of law that fails to take account of these finality interests would do more than subvert the criminal process itself. It would also seriously distort the very limited resources society has allocated to the criminal process. While men languish in jail, not uncommonly for over a year, awaiting a first trial on their guilt or innocence, it is not easy to justify expending substantial quantities of the time and energies of judges, prosecutors, and defense lawyers litigating the validity under present law of criminal convictions that were perfectly free from error when made final. See Friendly, supra, at 148-149. This drain on society’s resources is compounded by the fact that issuance of the habeas writ compels a State that wishes to continue enforcing its laws against the successful petitioner to relitigate facts buried in the remote past through presentation of witnesses whose memories of the relevant events often have dimmed. This very act of trying stale facts may well, ironically, produce a second trial no more reliable as a matter of getting at the truth than the first. See Amsterdam, Search, Seizure, and Section 2255: A Comment, 112 U. Pa. L. Rev. 378, 384 (1964).
In sum, while the case for continually inquiring into the current constitutional validity of criminal convictions *692on collateral attack is not an insubstantial one, it is by no means overwhelming. Most interests such a doctrine would serve will be adequately protected by the current rule that all constitutional errors not waived or harmless are correctible on habeas and by defining such errors according to the law in effect when a conviction became final. Those interests not served by this intermediate position are, in my view, largely overridden by the interests in finality.
Although not necessary to the resolution of either of the two collateral cases now here, for sake of completeness I venture to add that I would make two exceptions to this general principle. First, the above discussion is written only with new “procedural due process” rules in mind, that is, those applications of the Constitution that forbid the Government to utilize certain techniques or processes in enforcing concededly valid societal proscriptions on individual behavior.6 New “substantive due process” rules, that is, those that place, as a matter of constitutional interpretation, certain kinds of primary, private individual conduct beyond the power of the criminal law-making authority to proscribe,7 must, in my view, be placed on a different footing. As I noted above, the writ has historically *693been available for attacking convictions on such grounds.8 This, I believe, is because it represents the clearest instance where finality interests should yield. There is little societal interest in permitting the criminal process to rest at a point where it ought properly never to repose. Moreover, issuance of the writ on substantive due process grounds entails none of the adverse collateral consequences of retrial I have described above. Thus, the obvious interest in freeing individuals from punishment for conduct that is constitutionally protected seems to me sufficiently substantial to justify applying current notions of substantive due process to petitions for habeas corpus. See generally Part II of my opinion for the Court in United States v. United States Coin & Currency, post, p. 722.
Secondly, I think the writ ought always to lie for claims of nonobservance of those procedures that, as so aptly described by Mr. Justice Cardozo in Palko v. Connecticut, 302 U. S. 319, 325 (1937), are “implicit in the concept of ordered liberty.” Typically, it should be the case that any conviction free from federal constitutional error at the time it became final, will be found, upon reflection, to have been fundamentally fair and conducted under those procedures essential to the substance of a full hearing. However, in some situations it might be that time and growth in social capacity, as well as judicial perceptions of what we can rightly demand of the adjudicatory process, will properly alter our understanding of the bedrock procedural elements that must be found to vitiate the fairness of a particular conviction. *694For example, such, in my view, is the case with the right to counsel at trial now held a necessary condition precedent to any conviction for a serious crime. See my separate opinion in Gideon v. Wainwright, 372 U. S. 335, 349 (1963), where I concurred in conferring this right on a state prisoner, seeking state habeas corpus, on the grounds that this “new” rule was mandated by Palko. Hence, I would continue to apply Gideon itself on habeas, even to convictions made final before that decision was rendered. Other possible exceptions to the finality rule I would leave to be worked out in the context of actual cases brought before us that raise the issue.
Subsequent reflection upon what I wrote in Desist, where I undertook to expose in a preliminary way some of the considerations I thought ought to govern the problem of deciding which, if any, new constitutional rules should be held cognizable in habeas proceedings, leads me to these additional observations. There I tentatively suggested we might apply those new rules that “significantly improve the pre-existing fact-finding procedures” mandated by the Federal Constitution. 394 U. S., at 262. Cf. Mishkin, The Supreme Court 1964 Term — Foreword: The High Court, The Great Writ, and the Due Process of Time and Law, 79 Harv. L. Rev. 56, 77-101 (1965). As indicated above, I am now persuaded that those new rules cognizable on habeas ought to be defined, not by the “truth-determining” test, but by the Palko test. My reasons are several. First, adherence to precedent, particularly Kaufman v. United States, must ineluctably lead one to the conclusion that it is not a principal purpose of the writ to inquire whether a criminal convict did in fact commit the deed alleged. Additionally, recent decisions of this Court, e. g., Coleman v. Alabama, 399 U. S. 1 (1970), have revealed just how marginally effective are some new rules purportedly aimed at improving the fact-*695finding process. I cannot believe that the interest in finality is always outweighed by the interests protected in cases like Coleman. Cf. Spencer v. Texas, 385 U. S., at 583 (Warren, C. J., concurring and dissenting). I believe Palko more correctly marks the tipping point of finality interests, not only in terms of divining which new rules should apply on habeas, but also in its reminder that a particular rule may be more or less crucial to the fairness of a case depending on its own factual setting. Finally, I find inherently intractable the purported distinction between those new rules that are designed to improve the factfinding process and those designed principally to further other values. For a perfect example, note the plurality’s difficulty today in explaining, on that basis, retroactivity decisions such as Johnson v. New Jersey, 384 U. S. 719 (1966); Stovall v. Denno, 388 U. S. 293 (1967); and DeStefano v. Woods, 392 U. S. 631 (1968). Williams v. United States, ante, at 655-656, n. 7.
Secondly, in Desist I went to some lengths to point out the inevitable difficulties that will arise in attempting “to determine whether a particular decision has really announced a 'new’ rule at all or whether it has simply applied a well-established constitutional principle to govern a case which is closely analogous to those which have been previously considered in the prior case law.” 394 U. S., at 263. See generally id., at 263-269. I remain fully cognizant of these problems and realize they will produce some difficulties in administering the writ, but believe they would be greatly ameliorated by adequate recognition of the principle of finality in the operation of the criminal process.
Ill
I realize, of course, that this opinion, which is already unfortunately lengthy, has thus far not been directly responsive to the plurality opinions announced in these *696cases. Essentially this is because I do not perceive the issues raised by these cases from the same perspective as my Brethren. Certain aspects of the other opinions announced today do, however, fairly call for a response both because they contain some small seeds of a challenge to what I have said above and because I think, with respect, that what is written today by some of those who would give virtually unlimited sweep to this “retro-activity” doctrine strikingly illuminates the faulty premises of the thinking in this entire field.
In the plurality opinions in Williams and Elkmich, and Mackey the only challenge I perceive to my views is the single assertion that my analysis is untenable because unsupported by precedent. Williams v. United States, ante, at 651-652. Truly, this is a remarkable claim. For Linkletter v. Walker, supra, the wellspring of the current retroactivity doctrine, took as its point of departure the very distinction between direct review and collateral attack which I have argued is crucial to any analysis in this field, a distinction which the Court now firmly discards.
Further, as the dissenting opinion in United States v. United States Coin & Currency, post, at 735, points out, in an analogous situation, the legislative repeal of a criminal statute, “the judge-made rule was that those whose convictions had been finally affirmed when repeal took place received no benefit from the new rule; but repeal of a statute abated pending prosecutions and required reversal of convictions still on appeal when the law was changed.” In other words, the precise distinction I have urged between direct review and collateral attack, based not on the nature of the act of changing the law or of the new law thus pronounced but, instead, on the nature of the adjudicatory context in which the claim of legal error was presented has consistently *697been the model for the judicial process. Indeed, it would seem that the only precedential support for the position that prevails today is that conflicting and confusing flurry of “retroactivity” opinions that commenced less than five years ago with Johnson v. New Jersey, 384 U. S. 719 (1966).
Other aspects of the dissent in Coin & Currency, supra, might, it seems to me, be construed as a further challenge to the views I have expressed here since that opinion is subscribed to by a majority of those members of the Court who have determined that, for purposes of deciding whether new search and seizure rules apply to subsequent cases arising in federal courts, the process invoked by the litigants is irrelevant. In any event, I find the implications of the analysis underlying that dissent startling. For example, that Congress currently provides that statutory repeal shall not abate pending prosecutions or require reversal of nonfinal convictions seems to me a singularly unhelpful bit of information. We sit as a court of law, not a council of revision. Our powers of judicial review are judicial, not legislative, in nature. The assertion that this evidence is relevant data for resolving the problems at hand serves at best only to make explicit that which I have attempted to demonstrate in Part I of this opinion — that the retroactivity analysis currently ascendant in this Court proceeds on the false and unacceptable premise that constitutional interpretation is not purely a judicial, but, rather, something akin to a legislative, process. If, in fact, that premise is true we ought not to be writing retroactivity opinions but instead relinquishing some of our powers of judicial review.
The dissenting opinion attempts to palliate its invocation of the legislative process by alternately suggesting that the typical statutory rule is, because widespread, part *698of our fabric of “positive law” and the issue, therefore, is whether this Court should carry this policy over to the realm of constitutional interpretation. Three cases are cited that allegedly reveal we are not foreclosed from taking this course. The short answer to all this remains the same: the distinction between judicial and legislative power is equally woven deeply into the fabric of our positive law. So, too, is the notion that this Court definitively interprets the Constitution only because its role as a court of law requires it to do so. It is not surprising, then, to discover upon closer analysis that the cited cases do not bear the heavy weight placed on them. Gelpcke v. City of Dubuque, 1 Wall. 175 (1864), holds only that state courts may be compelled in some situations by particular provisions of the Federal Constitution to apply certain new rules prospectively only. No such claim has ever been made about these new constitutional rules of criminal procedure. Great Northern R. Co. v. Sunburst Oil & Refining Co., 287 U. S. 358 (1932), merely holds that the Federal Constitution imposes no barrier to a state court’s decision to apply a new state common-law rule prospectively only. Is it not sufficient answer to the dissenters’ final assertion of precedential support to point out that Chicot County Drainage District v. Baxter State Bank, 308 U. S. 371 (1940), was a collateral attack on a civil judgment already otherwise final and entitled to res judicata effect? And, further, that it was written by the same Chief Justice, Hughes, who had held six years earlier in United States v. Chambers, 291 U. S. 217 (1934), that repeal of the Eighteenth Amendment abated all prosecutions begun, and required reversal on direct review of all convictions obtained, under statutes dependent for their constitutionality on the repealed amendment, yet did not affect final convictions so obtained?
*699IV
Because my comprehension of the relevant issues diverges so substantially from that of the Court it is necessary for me to discuss separately my view as to the proper disposition of each of these three cases.
A
Williams v. United States (direct review). As this case is here on direct review, I would apply to its resolution the rule enunciated in Chimel v. California, 395 U. S. 752. The plurality correctly describes the salient facts in this case at n. 2 of its opinion, ante, at 650-651, and I agree they plainly reveal a violation of Chimel. Indeed, the Ninth Circuit panel below, although it held Chimel nonretroactive, explicitly found the search here involved inconsistent with the dictates of Chimel. 418 F. 2d 159, 161 (CA9 1969). Consequently, I would reverse the judgment below and remand with instructions to vacate the judgment of conviction.
B
Elkanich v. United States (collateral review). I agree, but for wholly different reasons, with the Court’s view, expressed in n. 2 of its opinion, ante, at 651, that we need not evaluate the search of Elkanich’s apartment in light of the precepts of Chimel. His conviction became final five years prior to Chimel’s promulgation, and prevailing law at that time certainly validated the search here involved. See United States v. Rabinowitz, 339 U. S. 56 (1950), and Harris v. United States, 331 U. S. 145 (1947). An appraisal of the facts surrounding this search leads me quite easily to conclude that the procedures used in obtaining this conviction were not so fundamentally devoid of the necessary elements of pro*700cedural due process as to require upsetting this conviction in spite of the fact that it was perfectly lawful when made final. The agents here clearly had probable cause to arrest petitioner, were not undertaking a fishing expedition for any evidence they might find but, rather, were looking for specific items that they had reason to believe might be concealed in various places around the premises and, indeed, generally limited their search to areas indicated by petitioner. I would affirm the judgment below.
C
Mackey v. United States (collateral review). Petitioner in this case seeks relief from confinement by way of habeas. At his trial for evading payment of income taxes, part of the Government’s case in chief consisted of the introduction of 60 wagering excise tax returns. At the time his conviction became final in 1965, the introduction of these statements would have been permissible under the authority of United States v. Kahriger, 345 U. S. 22 (1953). I find it unnecessary to inquire whether it inevitably follows from the new rule enunciated in Marchetti and Grosso that such a procedure would today be held an unacceptable abridgment of petitioner’s Fifth Amendment right to be free of compulsory self-incrimination. For, even assuming the latter cases, if applicable, would produce a different result, I cannot conclude that this change in the law would be sufficient to entitle petitioner to the issuance of a writ of habeas corpus.
Mackey is not asserting that the conduct for which he is being punished, evading payment of his federal income taxes, has been held to be constitutionally immune from punishment. In this regard, Mackey’s claim differs from that raised by the respondent in Coin & Currency, also decided today, where Marchetti and Grosso do operate to render Congress powerless to punish *701the conduct there at issue. Instead, Mackey's claim is that the procedures utilized in procuring his conviction were vitiated by the Marchetti and Grosso decisions. Since matters of procedure rather than substance are involved, see Part II of this opinion, I would apply to the resolution of this habeas petition the law in effect at the time Mackey’s conviction became final, absent a showing that the procedures employed were fundamentally unfair. While Kahriger did indeed, in my judgment, rest upon an “excessively narrow” view of the scope of the privilege against self-incrimination, I cannot say that hindsight reveals that judgment to have been so grossly erroneous as to amount to the perpetration of an inexcusable inequity against Mackey in these circumstances. Despite our rejection of it as a matter of Fifth Amendment policy, the prior justification of the Government’s activity in this area — that persons affected could avoid incrimination by ceasing to engage in illegal activities — is not without some force.
Although the question is, for me, not free of difficulty, I would affirm the judgment below for the reasons stated above.
V
In conclusion, the Court in deciding these cases seems largely to have forgotten the limitations that accompany its functions as a court of law. For the retroactivity doctrine announced today bespeaks more considerations of policy than of legal principle. Treating direct and collateral review as if they were of one piece seems to me faulty analysis, ignoring, as it does, the jurisprudential considerations that differentiate the two kinds of adjudicatory functions. As a court of law we have no right on direct review to treat one case differently from another with respect to constitutional provisions applicable to both. As regards cases coming here on collateral review, the problem of retroactivity is in truth *702none other than one of resettling the limits of the reach of the Great Writ, which under the recent decisions of this Court has been given almost boundless sweep.9 Until the Court is prepared to do this I can see no really satisfactory solution to the retroactivity problem. Meanwhile, I very much regret to see the existing free-wheeling approach to that problem now rewritten into the jurisprudence of this Court.
I would affirm the judgments in Nos. 36 and 82 and reverse the judgment in No. 81 upon the premises discussed in this opinion.
I realize, of course, that state prisoners are entitled to seek release via habeas corpus under 28 U. S. C. § 2241, while federal prisoners technically utilize what is denominated a motion to vacate judgment under 28 U. S. C. § 2255. However, our cases make these remedies virtually congruent and the purpose of substituting a motion to vacate for the traditional habeas action in the federal system was simply to alter one minor jurisdictional basis for the writ. See United States v. Hayman, 342 U. S. 205 (1952). As I do not propose to make any distinction, for retroactivity purposes, *682between state and federal prisoners seeking collateral relief, I shall refer throughout this opinion to both procedures as the writ of habeas corpus, and cases before us involving such judgments as cases here on collateral review.
For example, we have more than once in recent years had before us a libel case in which a party was allegedly libeled and brought suit for redress prior to this Court’s decision in New York Times Co. v. Sullivan, 376 U. S. 264 (1964), where we announced a new constitutional rule governing liability in libel suits brought by public officials. Yet no one connected with such cases has ever been heard to do so much as hint that the New York Times rule is not applicable because the conduct complained of occurred or the suit was brought before this new rule was promulgated. See, e. g., Rosenblatt v. Baer, 383 U. S. 75 (1966). Cf. Thorpe v. Housing Authority, 393 U. S. 268, 281-283 (1969).
Conversely, is it not perfectly clear that, had such a party procured and collected a final damage award prior to New York Times, the defendant could not have urged that the case be reopened solely because of our subsequent decision in that case? Absent proof of fraud or want of jurisdiction in the trial court that judgment would be res judicata and entitled to full faith and credit throughout the land.
This is not to suggest that civil and criminal collateral attack ought necessarily to be precisely congruent in the federal system. But certainly it illustrates that the law has always perceived collateral attack as a problem quite different from direct appeal.
Professor Mishkin has pointed out that “prior to Linkletter, the criteria applied in federal habeas corpus proceedings were uniformly the constitutional standards in effect at the time of those proceedings, regardless of when the conviction was actually entered.” Mishkin, The Supreme Court 1964 Term — Foreword: The High Court, the Great Writ, and the Due Process of Time and Law, 79 Harv. L. Rev. 56, 78 (1965). See also, e. g., Note, Developments in the Law — Federal Habeas Corpus, 83 Harv. L. Rev. 1040, 1151, 1153 (1970); Sanders v. United States, 373 U. S. 1, 17 (1963); Miller v. Gladden, 341 F. 2d 972, 975 (CA9 1965).
For a counter-example, see Milton v. Wainwright, 306 F. Supp. 929 (SD Fla.1969), where a district judge adjudicating a habeas petition declined to consider any of this Court’s decisions relating to involuntary confessions that postdated 1958, the time at which the petitioner’s murder conviction became final. See also n. 4, infra.
Arguably, Reck v. Pate, 367 U. S. 433 (1961), tacitly holds that habeas petitions must be judged in accordance with current law. The Court there directed the issuance of the writ on the ground that petitioner’s conviction, which became final in 1936, had been procured by the introduction into evidence of an illegally obtained confession, relying heavily on cases decided by this Court subsequent to 1936. The District Court, in denying relief, had clearly held that the admissibility of his confession was to be judged by standards prevailing in 1936. United States ex rel. Reck v. Ragen, 172 F. Supp. 734, 745-746 (ND Ill. 1959). However, this choice of law problem was not expressly adverted to and the case arose before this Court produced the recent enlargement of new constitutional rules of criminal procedure.
In 1966, Congress amended the habeas statutes to deal with this Court’s discussion in Sanders v. United States, 373 U. S. 1 (1963), of res judicata principles as they apply to habeas corpus. One subsection of that new statute provides:
“In a habeas corpus proceeding brought in behalf of a person in custody pursuant to the judgment of a State court, a prior judgment of the Supreme Court of the United States on an appeal or review ... of the decision of such State court, shall be conclusive as to all issues of fact or law . . . actually adjudicated by the Supreme Court therein, unless the applicant . . . shall plead and the court shall find the existence of a material and controlling *688fact which did not appear in the record of the proceeding in the Supreme Court [and could not have been put in by exercising due diligence].” 28 U. S. C. §2244 (c) (1964 ed., Supp. V) (emphasis added).
Unless one is to read “fact” as including a change in the law, it would seem that Congress has provided in these circumstances for finality as to legal determinations. That “fact” is properly read narrowly seems the better view in light of subsections (a) and (b) which permit a subsequent habeas petition (where there was no Supreme Court review) if it presents a “new ground” or “a factual or other ground not adjudicated on thé [prior] hearing.” Although the legislative history is extremely sparse, it fully supports this reading. Both the House and Senate committee reports accompanying these amendments stated that the purpose of the reformulation of § 2244 was to introduce a greater measure of finality into the law by providing for a qualified application of the res judicata concept. See H. R. Rep. No. 1892, 89th Cong., 2d Sess., 3, 8 (1966); S. Rep. No. 1797, 89th Cong., 2d Sess., 2 (1966). There was no relevant floor debate on these amendments.
Nor do I think the converse inference can properly be drawn that, if Congress provided legal finality for those prisoners whose convictions had been affirmed by us, it intentionally determined that other convicts should be able to avail themselves of all new constitutional rules on habeas. The language of subsections (a) and (b) certainly does not compel such a conclusion. The congressional committee reports neither state nor fairly imply that these amendments were designed to achieve the maximum feasible or desirable finality in habeas proceedings. Most important, it is difficult to imagine what would be the rationale for such a distinction merely between those who have and have not, at some time in the remote past, had full review of their cases in this Court.
I have in mind, of course, decisions such as Gideon v. Wainwright, 372 U. S. 335 (1963); Mapp v. Ohio, 367 U. S. 643 (1961); Miranda v. Arizona, 384 U. S. 436 (1966); Chimel v. California, 395 U. S. 752 (1969).
For example, Street v. New York, 394 U. S. 576 (1969); Stanley v. Georgia, 394 U. S. 557 (1969); Griswold v. Connecticut, 381 U. S. 479 (1965); Loving v. Virginia, 388 U. S. 1 (1967). Some rules may have both procedural and substantive ramifications, as I have used those terms here. See, e. g., my discussion, in Part IV-C of this opinion of the divergent ways Marchetti v. United States, 390 U. S. 39 (1968), and Grosso v. United States, 390 U. S. 62 (1968), bear on the problems raised by today’s Fifth Amendment cases.
See, e. g., Ex parte Siebold, 100 U. S. 371 (1880); Crowley v. Christensen, 137 U. S. 86 (1890); Yick Wo v. Hopkins, 118 U. S. 356 (1886). And see cases collected in Amsterdam, Search, Seizure, and Section 2255: A Comment, 112 U. Pa. L. Fev. 378, 384 n. 30 (1964), and the discussion therein of the finality implications such instances present.
For example, though correct in its result, I am now of the view that Linkletter would have been better decided had it simply held that federal habeas corpus does not lie for claimed errors in the introduction of illegally seized evidence.