JUSTICE O'CONNOR delivered the opinion of the Court.

This case requires us to consider the proper standards for judging a criminal defendant's contention that the Constitution requires a conviction or death sentence to be set aside because counsel's assistance at the trial or sentencing was ineffective.

I A

During a 10-day period in September 1976, respondent planned and committed three groups of crimes, which included three brutal stabbing murders, torture, kidnaping, severe assaults, attempted murders, attempted extortion, and theft. After his two accomplices were arrested, respondent surrendered to police and voluntarily gave a lengthy statement confessing to the third of the criminal episodes. The State of Florida indicted respondent for kidnaping and murder and appointed an experienced criminal lawyer to represent him.

Counsel actively pursued pretrial motions and discovery. He cut his efforts short, however, and he experienced a sense of hopelessness about the case, when he learned that, against his specific advice, respondent had also confessed to the first two murders. By the date set for trial, respondent was subject to indictment for three counts of first-degree murder and multiple counts of robbery, kidnaping for ransom, breaking and entering and assault, attempted murder, and conspiracy to commit robbery. Respondent waived his right to a jury trial, again acting against counsel's advice, and pleaded guilty to all charges, including the three capital murder charges.

In the plea colloquy, respondent told the trial judge that, although he had committed a string of burglaries, he had no significant prior criminal record and that at the time of his criminal spree he was under extreme stress caused by his inability to support his family. App. 50-53. He also stated, however, that he accepted responsibility for the crimes. E. g., id., at 54, 57. The trial judge told respondent that he had "a great deal of respect for people who are willing to step forward and admit their responsibility" but that he was making no statement at all about his likely sentencing decision. Id., at 62.

Counsel advised respondent to invoke his right under Florida law to an advisory jury at his capital sentencing hearing. Respondent rejected the advice and waived the right. He chose instead to be sentenced by the trial judge without a jury recommendation.

In preparing for the sentencing hearing, counsel spoke with respondent about his background. He also spoke on the telephone with respondent's wife and mother, though he did not follow up on the one unsuccessful effort to meet with them. He did not otherwise seek out character witnesses for respondent. App. to Pet. for Cert. A265. Nor did he request a psychiatric examination, since his conversations with his client gave no indication that respondent had psychological problems. Id., at A266.

Counsel decided not to present and hence not to look further for evidence concerning respondent's character and emotional state. That decision reflected trial counsel's sense of hopelessness about overcoming the evidentiary effect of respondent's confessions to the gruesome crimes. See id., at A282. It also reflected the judgment that it was advisable to rely on the plea colloquy for evidence about respondent's background and about his claim of emotional stress: the plea colloquy communicated sufficient information about these subjects, and by forgoing the opportunity to present new evidence on these subjects, counsel prevented the State from cross-examining respondent on his claim and from putting on psychiatric evidence of its own. Id., at A223-A225.

Counsel also excluded from the sentencing hearing other evidence he thought was potentially damaging. He successfully moved to exclude respondent's "rap sheet." Id., at A227; App. 311. Because he judged that a presentence report might prove more detrimental than helpful, as it would have included respondent's criminal history and thereby would have undermined the claim of no significant history of criminal activity, he did not request that one be prepared. App. to Pet. for Cert. A227-A228, A265-A266.

At the sentencing hearing, counsel's strategy was based primarily on the trial judge's remarks at the plea colloquy as well as on his reputation as a sentencing judge who thought it important for a convicted defendant to own up to his crime. Counsel argued that respondent's remorse and acceptance of responsibility justified sparing him from the death penalty. Id., at A265-A266. Counsel also argued that respondent had no history of criminal activity and that respondent committed the crimes under extreme mental or emotional disturbance, thus coming within the statutory list of mitigating circumstances. He further argued that respondent should be spared death because he had surrendered, confessed, and offered to testify against a codefendant and because respondent was fundamentally a good person who had briefly gone badly wrong in extremely stressful circumstances. The State put on evidence and witnesses largely for the purpose of describing the details of the crimes. Counsel did not cross-examine the medical experts who testified about the manner of death of respondent's victims.

The trial judge found several aggravating circumstances with respect to each of the three murders. He found that all three murders were especially heinous, atrocious, and cruel, all involving repeated stabbings. All three murders were committed in the course of at least one other dangerous and violent felony, and since all involved robbery, the murders were for pecuniary gain. All three murders were committed to avoid arrest for the accompanying crimes and to hinder law enforcement. In the course of one of the murders, respondent knowingly subjected numerous persons to a grave risk of death by deliberately stabbing and shooting the murder victim's sisters-in-law, who sustained severe — in one case, ultimately fatal — injuries.

With respect to mitigating circumstances, the trial judge made the same findings for all three capital murders. First, although there was no admitted evidence of prior convictions, respondent had stated that he had engaged in a course of stealing. In any case, even if respondent had no significant history of criminal activity, the aggravating circumstances "would still clearly far outweigh" that mitigating factor. Second, the judge found that, during all three crimes, respondent was not suffering from extreme mental or emotional disturbance and could appreciate the criminality of his acts. Third, none of the victims was a participant in, or consented to, respondent's conduct. Fourth, respondent's participation in the crimes was neither minor nor the result of duress or domination by an accomplice. Finally, respondent's age (26) could not be considered a factor in mitigation, especially when viewed in light of respondent's planning of the crimes and disposition of the proceeds of the various accompanying thefts.

In short, the trial judge found numerous aggravating circumstances and no (or a single comparatively insignificant) mitigating circumstance. With respect to each of the three convictions for capital murder, the trial judge concluded: "A careful consideration of all matters presented to the court impels the conclusion that there are insufficient mitigating circumstances . . . to outweigh the aggravating circumstances." See Washington v. State, 362 So.2d 658, 663-664 (Fla. 1978) (quoting trial court findings), cert. denied, 441 U.S. 937 (1979). He therefore sentenced respondent to death on each of the three counts of murder and to prison terms for the other crimes. The Florida Supreme Court upheld the convictions and sentences on direct appeal.

B

Respondent subsequently sought collateral relief in state court on numerous grounds, among them that counsel had rendered ineffective assistance at the sentencing proceeding. Respondent challenged counsel's assistance in six respects. He asserted that counsel was ineffective because he failed to move for a continuance to prepare for sentencing, to request a psychiatric report, to investigate and present character witnesses, to seek a presentence investigation report, to present meaningful arguments to the sentencing judge, and to investigate the medical examiner's reports or cross-examine the medical experts. In support of the claim, respondent submitted 14 affidavits from friends, neighbors, and relatives stating that they would have testified if asked to do so. He also submitted one psychiatric report and one psychological report stating that respondent, though not under the influence of extreme mental or emotional disturbance, was "chronically frustrated and depressed because of his economic dilemma" at the time of his crimes. App. 7; see also id., at 14.

The trial court denied relief without an evidentiary hearing, finding that the record evidence conclusively showed that the ineffectiveness claim was meritless. App. to Pet. for Cert. A206-A243. Four of the assertedly prejudicial errors required little discussion. First, there were no grounds to request a continuance, so there was no error in not requesting one when respondent pleaded guilty. Id., at A218-A220. Second, failure to request a presentence investigation was not a serious error because the trial judge had discretion not to grant such a request and because any presentence investigation would have resulted in admission of respondent's "rap sheet" and thus would have undermined his assertion of no significant history of criminal activity. Id., at A226-A228. Third, the argument and memorandum given to the sentencing judge were "admirable" in light of the overwhelming aggravating circumstances and absence of mitigating circumstances. Id., at A228. Fourth, there was no error in failure to examine the medical examiner's reports or to cross-examine the medical witnesses testifying on the manner of death of respondent's victims, since respondent admitted that the victims died in the ways shown by the unchallenged medical evidence. Id., at A229.

The trial court dealt at greater length with the two other bases for the ineffectiveness claim. The court pointed out that a psychiatric examination of respondent was conducted by state order soon after respondent's initial arraignment. That report states that there was no indication of major mental illness at the time of the crimes. Moreover, both the reports submitted in the collateral proceeding state that, although respondent was "chronically frustrated and depressed because of his economic dilemma," he was not under the influence of extreme mental or emotional disturbance. All three reports thus directly undermine the contention made at the sentencing hearing that respondent was suffering from extreme mental or emotional disturbance during his crime spree. Accordingly, counsel could reasonably decide not to seek psychiatric reports; indeed, by relying solely on the plea colloquy to support the emotional disturbance contention, counsel denied the State an opportunity to rebut his claim with psychiatric testimony. In any event, the aggravating circumstances were so overwhelming that no substantial prejudice resulted from the absence at sentencing of the psychiatric evidence offered in the collateral attack.

The court rejected the challenge to counsel's failure to develop and to present character evidence for much the same reasons. The affidavits submitted in the collateral proceeding showed nothing more than that certain persons would have testified that respondent was basically a good person who was worried about his family's financial problems. Respondent himself had already testified along those lines at the plea colloquy. Moreover, respondent's admission of a course of stealing rebutted many of the factual allegations in the affidavits. For those reasons, and because the sentencing judge had stated that the death sentence would be appropriate even if respondent had no significant prior criminal history, no substantial prejudice resulted from the absence at sentencing of the character evidence offered in the collateral attack.

Applying the standard for ineffectiveness claims articulated by the Florida Supreme Court in Knight v. State, 394 So.2d 997 (1981), the trial court concluded that respondent had not shown that counsel's assistance reflected any substantial and serious deficiency measurably below that of competent counsel that was likely to have affected the outcome of the sentencing proceeding. The court specifically found: "[A]s a matter of law, the record affirmatively demonstrates beyond any doubt that even if [counsel] had done each of the . . . things [that respondent alleged counsel had failed to do] at the time of sentencing, there is not even the remotest chance that the outcome would have been any different. The plain fact is that the aggravating circumstances proved in this case were completely overwhelming . . . ." App. to Pet. for Cert. A230.

The Florida Supreme Court affirmed the denial of relief. Washington v. State, 397 So.2d 285 (1981). For essentially the reasons given by the trial court, the State Supreme Court concluded that respondent had failed to make out a prima facie case of either "substantial deficiency or possible prejudice" and, indeed, had "failed to such a degree that we believe, to the point of a moral certainty, that he is entitled to no relief . . . ." Id., at 287. Respondent's claims were "shown conclusively to be without merit so as to obviate the need for an evidentiary hearing." Id., at 286.

C

Respondent next filed a petition for a writ of habeas corpus in the United States District Court for the Southern District of Florida. He advanced numerous grounds for relief, among them ineffective assistance of counsel based on the same errors, except for the failure to move for a continuance, as those he had identified in state court. The District Court held an evidentiary hearing to inquire into trial counsel's efforts to investigate and to present mitigating circumstances. Respondent offered the affidavits and reports he had submitted in the state collateral proceedings; he also called his trial counsel to testify. The State of Florida, over respondent's objection, called the trial judge to testify.

The District Court disputed none of the state court factual findings concerning trial counsel's assistance and made findings of its own that are consistent with the state court findings. The account of trial counsel's actions and decisions given above reflects the combined findings. On the legal issue of ineffectiveness, the District Court concluded that, although trial counsel made errors in judgment in failing to investigate nonstatutory mitigating evidence further than he did, no prejudice to respondent's sentence resulted from any such error in judgment. Relying in part on the trial judge's testimony but also on the same factors that led the state courts to find no prejudice, the District Court concluded that "there does not appear to be a likelihood, or even a significant possibility," that any errors of trial counsel had affected the outcome of the sentencing proceeding. App. to Pet. for Cert. A285-A286. The District Court went on to reject all of respondent's other grounds for relief, including one not exhausted in state court, which the District Court considered because, among other reasons, the State urged its consideration. Id., at A286-A292. The court accordingly denied the petition for a writ of habeas corpus.

On appeal, a panel of the United States Court of Appeals for the Fifth Circuit affirmed in part, vacated in part, and remanded with instructions to apply to the particular facts the framework for analyzing ineffectiveness claims that it developed in its opinion. 673 F.2d 879 (1982). The panel decision was itself vacated when Unit B of the former Fifth Circuit, now the Eleventh Circuit, decided to rehear the case en banc. 679 F.2d 23 (1982). The full Court of Appeals developed its own framework for analyzing ineffective assistance claims and reversed the judgment of the District Court and remanded the case for new factfinding under the newly announced standards. 693 F.2d 1243 (1982).

The court noted at the outset that, because respondent had raised an unexhausted claim at his evidentiary hearing in the District Court, the habeas petition might be characterized as a mixed petition subject to the rule of Rose v. Lundy, 455 U.S. 509 (1982), requiring dismissal of the entire petition. The court held, however, that the exhaustion requirement is "a matter of comity rather than a matter of jurisdiction" and hence admitted of exceptions. The court agreed with the District Court that this case came within an exception to the mixed petition rule. 693 F.2d, at 1248, n. 7.

Turning to the merits, the Court of Appeals stated that the Sixth Amendment right to assistance of counsel accorded criminal defendants a right to "counsel reasonably likely to render and rendering reasonably effective assistance given the totality of the circumstances." Id., at 1250. The court remarked in passing that no special standard applies in capital cases such as the one before it: the punishment that a defendant faces is merely one of the circumstances to be considered in determining whether counsel was reasonably effective. Id., at 1250, n. 12. The court then addressed respondent's contention that his trial counsel's assistance was not reasonably effective because counsel breached his duty to investigate nonstatutory mitigating circumstances.

The court agreed that the Sixth Amendment imposes on counsel a duty to investigate, because reasonably effective assistance must be based on professional decisions and informed legal choices can be made only after investigation of options. The court observed that counsel's investigatory decisions must be assessed in light of the information known at the time of the decisions, not in hindsight, and that "[t]he amount of pretrial investigation that is reasonable defies precise measurement." Id., at 1251. Nevertheless, putting guilty-plea cases to one side, the court attempted to classify cases presenting issues concerning the scope of the duty to investigate before proceeding to trial.

If there is only one plausible line of defense, the court concluded, counsel must conduct a "reasonably substantial investigation" into that line of defense, since there can be no strategic choice that renders such an investigation unnecessary. Id., at 1252. The same duty exists if counsel relies at trial on only one line of defense, although others are available. In either case, the investigation need not be exhaustive. It must include "'an independent examination of the facts, circumstances, pleadings and laws involved.'" Id., at 1253 (quoting Rummel v. Estelle, 590 F.2d 103, 104 (CA5 1979)). The scope of the duty, however, depends on such facts as the strength of the government's case and the likelihood that pursuing certain leads may prove more harmful than helpful. 693 F.2d, at 1253, n. 16.

If there is more than one plausible line of defense, the court held, counsel should ideally investigate each line substantially before making a strategic choice about which lines to rely on at trial. If counsel conducts such substantial investigations, the strategic choices made as a result "will seldom if ever" be found wanting. Because advocacy is an art and not a science, and because the adversary system requires deference to counsel's informed decisions, strategic choices must be respected in these circumstances if they are based on professional judgment. Id., at 1254.

If counsel does not conduct a substantial investigation into each of several plausible lines of defense, assistance may nonetheless be effective. Counsel may not exclude certain lines of defense for other than strategic reasons. Id., at 1257-1258. Limitations of time and money, however, may force early strategic choices, often based solely on conversations with the defendant and a review of the prosecution's evidence. Those strategic choices about which lines of defense to pursue are owed deference commensurate with the reasonableness of the professional judgments on which they are based. Thus, "when counsel's assumptions are reasonable given the totality of the circumstances and when counsel's strategy represents a reasonable choice based upon those assumptions, counsel need not investigate lines of defense that he has chosen not to employ at trial." Id., at 1255 (footnote omitted). Among the factors relevant to deciding whether particular strategic choices are reasonable are the experience of the attorney, the inconsistency of unpursued and pursued lines of defense, and the potential for prejudice from taking an unpursued line of defense. Id., at 1256-1257, n. 23.

Having outlined the standards for judging whether defense counsel fulfilled the duty to investigate, the Court of Appeals turned its attention to the question of the prejudice to the defense that must be shown before counsel's errors justify reversal of the judgment. The court observed that only in cases of outright denial of counsel, of affirmative government interference in the representation process, or of inherently prejudicial conflicts of interest had this Court said that no special showing of prejudice need be made. Id., at 1258-1259. For cases of deficient performance by counsel, where the government is not directly responsible for the deficiencies and where evidence of deficiency may be more accessible to the defendant than to the prosecution, the defendant must show that counsel's errors "resulted in actual and substantial disadvantage to the course of his defense." Id., at 1262. This standard, the Court of Appeals reasoned, is compatible with the "cause and prejudice" standard for overcoming procedural defaults in federal collateral proceedings and discourages insubstantial claims by requiring more than a showing, which could virtually always be made, of some conceivable adverse effect on the defense from counsel's errors. The specified showing of prejudice would result in reversal of the judgment, the court concluded, unless the prosecution showed that the constitutionally deficient performance was, in light of all the evidence, harmless beyond a reasonable doubt. Id., at 1260-1262.

The Court of Appeals thus laid down the tests to be applied in the Eleventh Circuit in challenges to convictions on the ground of ineffectiveness of counsel. Although some of the judges of the court proposed different approaches to judging ineffectiveness claims either generally or when raised in federal habeas petitions from state prisoners, id., at 1264-1280 (opinion of Tjoflat, J.); id., at 1280 (opinion of Clark, J.); id., at 1285-1288 (opinion of Roney, J., joined by Fay and Hill, JJ.); id., at 1288-1291 (opinion of Hill, J.), and although some believed that no remand was necessary in this case, id., at 1281-1285 (opinion of Johnson, J., joined by Anderson, J.); id., at 1285-1288 (opinion of Roney, J., joined by Fay and Hill, JJ.); id., at 1288-1291 (opinion of Hill, J.), a majority of the judges of the en banc court agreed that the case should be remanded for application of the newly announced standards. Summarily rejecting respondent's claims other than ineffectiveness of counsel, the court accordingly reversed the judgment of the District Court and remanded the case. On remand, the court finally ruled, the state trial judge's testimony, though admissible "to the extent that it contains personal knowledge of historical facts or expert opinion," was not to be considered admitted into evidence to explain the judge's mental processes in reaching his sentencing decision. Id., at 1262-1263; see Fayerweather v. Ritch, 195 U.S. 276, 306-307 (1904).

D

Petitioners, who are officials of the State of Florida, filed a petition for a writ of certiorari seeking review of the decision of the Court of Appeals. The petition presents a type of Sixth Amendment claim that this Court has not previously considered in any generality. The Court has considered Sixth Amendment claims based on actual or constructive denial of the assistance of counsel altogether, as well as claims based on state interference with the ability of counsel to render effective assistance to the accused. E. g., United States v. Cronic, ante, p. 648. With the exception of Cuyler v. Sullivan, 446 U.S. 335 (1980), however, which involved a claim that counsel's assistance was rendered ineffective by a conflict of interest, the Court has never directly and fully addressed a claim of "actual ineffectiveness" of counsel's assistance in a case going to trial. Cf. United States v. Agurs, 427 U.S. 97, 102, n. 5 (1976).

In assessing attorney performance, all the Federal Courts of Appeals and all but a few state courts have now adopted the "reasonably effective assistance" standard in one formulation or another. See Trapnell v. United States, 725 F.2d 149, 151-152 (CA2 1983); App. B to Brief for United States in United States v. Cronic, O. T. 1983, No. 82-660, pp. 3a-6a; Sarno, Modern Status of Rules and Standards in State Courts as to Adequacy of Defense Counsel's Representation of Criminal Client, 2 A. L. R. 4th 99-157, §§ 7-10 (1980). Yet this Court has not had occasion squarely to decide whether that is the proper standard. With respect to the prejudice that a defendant must show from deficient attorney performance, the lower courts have adopted tests that purport to differ in more than formulation. See App. C to Brief for United States in United States v. Cronic, supra, at 7a-10a; Sarno, supra, at 83-99, § 6. In particular, the Court of Appeals in this case expressly rejected the prejudice standard articulated by Judge Leventhal in his plurality opinion in United States v. Decoster, 199 U.S.App.D.C. 359, 371, 374-375, 624 F.2d 196, 208, 211-212 (en banc), cert. denied, 444 U.S. 944 (1979), and adopted by the State of Florida in Knight v. State, 394 So.2d, at 1001, a standard that requires a showing that specified deficient conduct of counsel was likely to have affected the outcome of the proceeding. 693 F.2d, at 1261-1262.

For these reasons, we granted certiorari to consider the standards by which to judge a contention that the Constitution requires that a criminal judgment be overturned because of the actual ineffective assistance of counsel. 462 U.S. 1105 (1983). We agree with the Court of Appeals that the exhaustion rule requiring dismissal of mixed petitions, though to be strictly enforced, is not jurisdictional. See Rose v. Lundy, 455 U.S., at 515-520. We therefore address the merits of the constitutional issue.

II

In a long line of cases that includes Powell v. Alabama, 287 U.S. 45 (1932), Johnson v. Zerbst, 304 U.S. 458 (1938), and Gideon v. Wainwright, 372 U.S. 335 (1963), this Court has recognized that the Sixth Amendment right to counsel exists, and is needed, in order to protect the fundamental right to a fair trial. The Constitution guarantees a fair trial through the Due Process Clauses, but it defines the basic elements of a fair trial largely through the several provisions of the Sixth Amendment, including the Counsel Clause:

"In all criminal prosecutions, the accused shall enjoy the right to a speedy and public trial, by an impartial jury of the State and district wherein the crime shall have been committed, which district shall have been previously ascertained by law, and to be informed of the nature and cause of the accusation; to be confronted with the witnesses against him; to have compulsory process for obtaining witnesses in his favor, and to have the Assistance of Counsel for his defence."

Thus, a fair trial is one in which evidence subject to adversarial testing is presented to an impartial tribunal for resolution of issues defined in advance of the proceeding. The right to counsel plays a crucial role in the adversarial system embodied in the Sixth Amendment, since access to counsel's skill and knowledge is necessary to accord defendants the "ample opportunity to meet the case of the prosecution" to which they are entitled. Adams v. United States ex rel. McCann, 317 U.S. 269, 275, 276 (1942); see Powell v. Alabama, supra, at 68-69.

Because of the vital importance of counsel's assistance, this Court has held that, with certain exceptions, a person accused of a federal or state crime has the right to have counsel appointed if retained counsel cannot be obtained. See Argersinger v. Hamlin, 407 U.S. 25 (1972); Gideon v. Wainwright, supra; Johnson v. Zerbst, supra. That a person who happens to be a lawyer is present at trial alongside the accused, however, is not enough to satisfy the constitutional command. The Sixth Amendment recognizes the right to the assistance of counsel because it envisions counsel's playing a role that is critical to the ability of the adversarial system to produce just results. An accused is entitled to be assisted by an attorney, whether retained or appointed, who plays the role necessary to ensure that the trial is fair.

For that reason, the Court has recognized that "the right to counsel is the right to the effective assistance of counsel." McMann v. Richardson, 397 U.S. 759, 771, n. 14 (1970). Government violates the right to effective assistance when it interferes in certain ways with the ability of counsel to make independent decisions about how to conduct the defense. See, e. g., Geders v. United States, 425 U.S. 80 (1976) (bar on attorney-client consultation during overnight recess); Herring v. New York, 422 U.S. 853 (1975) (bar on summation at bench trial); Brooks v. Tennessee, 406 U.S. 605, 612-613 (1972) (requirement that defendant be first defense witness); Ferguson v. Georgia, 365 U.S. 570, 593-596 (1961) (bar on direct examination of defendant). Counsel, however, can also deprive a defendant of the right to effective assistance, simply by failing to render "adequate legal assistance," Cuyler v. Sullivan, 446 U.S., at 344. Id., at 345-350 (actual conflict of interest adversely affecting lawyer's performance renders assistance ineffective).

The Court has not elaborated on the meaning of the constitutional requirement of effective assistance in the latter class of cases — that is, those presenting claims of "actual ineffectiveness." In giving meaning to the requirement, however, we must take its purpose — to ensure a fair trial — as the guide. The benchmark for judging any claim of ineffectiveness must be whether counsel's conduct so undermined the proper functioning of the adversarial process that the trial cannot be relied on as having produced a just result.

The same principle applies to a capital sentencing proceeding such as that provided by Florida law. We need not consider the role of counsel in an ordinary sentencing, which may involve informal proceedings and standardless discretion in the sentencer, and hence may require a different approach to the definition of constitutionally effective assistance. A capital sentencing proceeding like the one involved in this case, however, is sufficiently like a trial in its adversarial format and in the existence of standards for decision, see Barclay v. Florida, 463 U.S. 939, 952-954 (1983); Bullington v. Missouri, 451 U.S. 430 (1981), that counsel's role in the proceeding is comparable to counsel's role at trial — to ensure that the adversarial testing process works to produce a just result under the standards governing decision. For purposes of describing counsel's duties, therefore, Florida's capital sentencing proceeding need not be distinguished from an ordinary trial.

III

A convicted defendant's claim that counsel's assistance was so defective as to require reversal of a conviction or death sentence has two components. First, the defendant must show that counsel's performance was deficient. This requires showing that counsel made errors so serious that counsel was not functioning as the "counsel" guaranteed the defendant by the Sixth Amendment. Second, the defendant must show that the deficient performance prejudiced the defense. This requires showing that counsel's errors were so serious as to deprive the defendant of a fair trial, a trial whose result is reliable. Unless a defendant makes both showings, it cannot be said that the conviction or death sentence resulted from a breakdown in the adversary process that renders the result unreliable.

A

As all the Federal Courts of Appeals have now held, the proper standard for attorney performance is that of reasonably effective assistance. See Trapnell v. United States, 725 F.2d, at 151-152. The Court indirectly recognized as much when it stated in McMann v. Richardson, supra, at 770, 771, that a guilty plea cannot be attacked as based on inadequate legal advice unless counsel was not "a reasonably competent attorney" and the advice was not "within the range of competence demanded of attorneys in criminal cases." See also Cuyler v. Sullivan, supra, at 344. When a convicted defendant complains of the ineffectiveness of counsel's assistance, the defendant must show that counsel's representation fell below an objective standard of reasonableness.

More specific guidelines are not appropriate. The Sixth Amendment refers simply to "counsel," not specifying particular requirements of effective assistance. It relies instead on the legal profession's maintenance of standards sufficient to justify the law's presumption that counsel will fulfill the role in the adversary process that the Amendment envisions. See Michel v. Louisiana, 350 U.S. 91, 100-101 (1955). The proper measure of attorney performance remains simply reasonableness under prevailing professional norms.

Representation of a criminal defendant entails certain basic duties. Counsel's function is to assist the defendant, and hence counsel owes the client a duty of loyalty, a duty to avoid conflicts of interest. See Cuyler v. Sullivan, supra, at 346. From counsel's function as assistant to the defendant derive the overarching duty to advocate the defendant's cause and the more particular duties to consult with the defendant on important decisions and to keep the defendant informed of important developments in the course of the prosecution. Counsel also has a duty to bring to bear such skill and knowledge as will render the trial a reliable adversarial testing process. See Powell v. Alabama, 287 U.S., at 68-69.

These basic duties neither exhaustively define the obligations of counsel nor form a checklist for judicial evaluation of attorney performance. In any case presenting an ineffectiveness claim, the performance inquiry must be whether counsel's assistance was reasonable considering all the circumstances. Prevailing norms of practice as reflected in American Bar Association standards and the like, e. g., ABA Standards for Criminal Justice 4-1.1 to 4-8.6 (2d ed. 1980) ("The Defense Function"), are guides to determining what is reasonable, but they are only guides. No particular set of detailed rules for counsel's conduct can satisfactorily take account of the variety of circumstances faced by defense counsel or the range of legitimate decisions regarding how best to represent a criminal defendant. Any such set of rules would interfere with the constitutionally protected independence of counsel and restrict the wide latitude counsel must have in making tactical decisions. See United States v. Decoster, 199 U.S. App. D.C., at 371, 624 F.2d, at 208. Indeed, the existence of detailed guidelines for representation could distract counsel from the overriding mission of vigorous advocacy of the defendant's cause. Moreover, the purpose of the effective assistance guarantee of the Sixth Amendment is not to improve the quality of legal representation, although that is a goal of considerable importance to the legal system. The purpose is simply to ensure that criminal defendants receive a fair trial.

Judicial scrutiny of counsel's performance must be highly deferential. It is all too tempting for a defendant to second-guess counsel's assistance after conviction or adverse sentence, and it is all too easy for a court, examining counsel's defense after it has proved unsuccessful, to conclude that a particular act or omission of counsel was unreasonable. Cf. Engle v. Isaac, 456 U.S. 107, 133-134 (1982). A fair assessment of attorney performance requires that every effort be made to eliminate the distorting effects of hindsight, to reconstruct the circumstances of counsel's challenged conduct, and to evaluate the conduct from counsel's perspective at the time. Because of the difficulties inherent in making the evaluation, a court must indulge a strong presumption that counsel's conduct falls within the wide range of reasonable professional assistance; that is, the defendant must overcome the presumption that, under the circumstances, the challenged action "might be considered sound trial strategy." See Michel v. Louisiana, supra, at 101. There are countless ways to provide effective assistance in any given case. Even the best criminal defense attorneys would not defend a particular client in the same way. See Goodpaster, The Trial for Life: Effective Assistance of Counsel in Death Penalty Cases, 58 N.Y. U. L. Rev. 299, 343 (1983).

The availability of intrusive post-trial inquiry into attorney performance or of detailed guidelines for its evaluation would encourage the proliferation of ineffectiveness challenges. Criminal trials resolved unfavorably to the defendant would increasingly come to be followed by a second trial, this one of counsel's unsuccessful defense. Counsel's performance and even willingness to serve could be adversely affected. Intensive scrutiny of counsel and rigid requirements for acceptable assistance could dampen the ardor and impair the independence of defense counsel, discourage the acceptance of assigned cases, and undermine the trust between attorney and client.

Thus, a court deciding an actual ineffectiveness claim must judge the reasonableness of counsel's challenged conduct on the facts of the particular case, viewed as of the time of counsel's conduct. A convicted defendant making a claim of ineffective assistance must identify the acts or omissions of counsel that are alleged not to have been the result of reasonable professional judgment. The court must then determine whether, in light of all the circumstances, the identified acts or omissions were outside the wide range of professionally competent assistance. In making that determination, the court should keep in mind that counsel's function, as elaborated in prevailing professional norms, is to make the adversarial testing process work in the particular case. At the same time, the court should recognize that counsel is strongly presumed to have rendered adequate assistance and made all significant decisions in the exercise of reasonable professional judgment.

These standards require no special amplification in order to define counsel's duty to investigate, the duty at issue in this case. As the Court of Appeals concluded, strategic choices made after thorough investigation of law and facts relevant to plausible options are virtually unchallengeable; and strategic choices made after less than complete investigation are reasonable precisely to the extent that reasonable professional judgments support the limitations on investigation. In other words, counsel has a duty to make reasonable investigations or to make a reasonable decision that makes particular investigations unnecessary. In any ineffectiveness case, a particular decision not to investigate must be directly assessed for reasonableness in all the circumstances, applying a heavy measure of deference to counsel's judgments.

The reasonableness of counsel's actions may be determined or substantially influenced by the defendant's own statements or actions. Counsel's actions are usually based, quite properly, on informed strategic choices made by the defendant and on information supplied by the defendant. In particular, what investigation decisions are reasonable depends critically on such information. For example, when the facts that support a certain potential line of defense are generally known to counsel because of what the defendant has said, the need for further investigation may be considerably diminished or eliminated altogether. And when a defendant has given counsel reason to believe that pursuing certain investigations would be fruitless or even harmful, counsel's failure to pursue those investigations may not later be challenged as unreasonable. In short, inquiry into counsel's conversations with the defendant may be critical to a proper assessment of counsel's investigation decisions, just as it may be critical to a proper assessment of counsel's other litigation decisions. See United States v. Decoster, supra, at 372-373, 624 F.2d, at 209-210.

B

An error by counsel, even if professionally unreasonable, does not warrant setting aside the judgment of a criminal proceeding if the error had no effect on the judgment. Cf. United States v. Morrison, 449 U.S. 361, 364-365 (1981). The purpose of the Sixth Amendment guarantee of counsel is to ensure that a defendant has the assistance necessary to justify reliance on the outcome of the proceeding. Accordingly, any deficiencies in counsel's performance must be prejudicial to the defense in order to constitute ineffective assistance under the Constitution.

In certain Sixth Amendment contexts, prejudice is presumed. Actual or constructive denial of the assistance of counsel altogether is legally presumed to result in prejudice. So are various kinds of state interference with counsel's assistance. See United States v. Cronic, ante, at 659, and n. 25. Prejudice in these circumstances is so likely that case-by-case inquiry into prejudice is not worth the cost. Ante, at 658. Moreover, such circumstances involve impairments of the Sixth Amendment right that are easy to identify and, for that reason and because the prosecution is directly responsible, easy for the government to prevent.

One type of actual ineffectiveness claim warrants a similar, though more limited, presumption of prejudice. In Cuyler v. Sullivan, 446 U.S., at 345-350, the Court held that prejudice is presumed when counsel is burdened by an actual conflict of interest. In those circumstances, counsel breaches the duty of loyalty, perhaps the most basic of counsel's duties. Moreover, it is difficult to measure the precise effect on the defense of representation corrupted by conflicting interests. Given the obligation of counsel to avoid conflicts of interest and the ability of trial courts to make early inquiry in certain situations likely to give rise to conflicts, see, e. g., Fed. Rule Crim. Proc. 44(c), it is reasonable for the criminal justice system to maintain a fairly rigid rule of presumed prejudice for conflicts of interest. Even so, the rule is not quite the per se rule of prejudice that exists for the Sixth Amendment claims mentioned above. Prejudice is presumed only if the defendant demonstrates that counsel "actively represented conflicting interests" and that "an actual conflict of interest adversely affected his lawyer's performance." Cuyler v. Sullivan, supra, at 350, 348 (footnote omitted).

Conflict of interest claims aside, actual ineffectiveness claims alleging a deficiency in attorney performance are subject to a general requirement that the defendant affirmatively prove prejudice. The government is not responsible for, and hence not able to prevent, attorney errors that will result in reversal of a conviction or sentence. Attorney errors come in an infinite variety and are as likely to be utterly harmless in a particular case as they are to be prejudicial. They cannot be classified according to likelihood of causing prejudice. Nor can they be defined with sufficient precision to inform defense attorneys correctly just what conduct to avoid. Representation is an art, and an act or omission that is unprofessional in one case may be sound or even brilliant in another. Even if a defendant shows that particular errors of counsel were unreasonable, therefore, the defendant must show that they actually had an adverse effect on the defense.

It is not enough for the defendant to show that the errors had some conceivable effect on the outcome of the proceeding. Virtually every act or omission of counsel would meet that test, cf. United States v. Valenzuela-Bernal, 458 U.S. 858, 866-867 (1982), and not every error that conceivably could have influenced the outcome undermines the reliability of the result of the proceeding. Respondent suggests requiring a showing that the errors "impaired the presentation of the defense." Brief for Respondent 58. That standard, however, provides no workable principle. Since any error, if it is indeed an error, "impairs" the presentation of the defense, the proposed standard is inadequate because it provides no way of deciding what impairments are sufficiently serious to warrant setting aside the outcome of the proceeding.

On the other hand, we believe that a defendant need not show that counsel's deficient conduct more likely than not altered the outcome in the case. This outcome-determinative standard has several strengths. It defines the relevant inquiry in a way familiar to courts, though the inquiry, as is inevitable, is anything but precise. The standard also reflects the profound importance of finality in criminal proceedings. Moreover, it comports with the widely used standard for assessing motions for new trial based on newly discovered evidence. See Brief for United States as Amicus Curiae 19-20, and nn. 10, 11. Nevertheless, the standard is not quite appropriate.

Even when the specified attorney error results in the omission of certain evidence, the newly discovered evidence standard is not an apt source from which to draw a prejudice standard for ineffectiveness claims. The high standard for newly discovered evidence claims presupposes that all the essential elements of a presumptively accurate and fair proceeding were present in the proceeding whose result is challenged. Cf. United States v. Johnson, 327 U.S. 106, 112 (1946). An ineffective assistance claim asserts the absence of one of the crucial assurances that the result of the proceeding is reliable, so finality concerns are somewhat weaker and the appropriate standard of prejudice should be somewhat lower. The result of a proceeding can be rendered unreliable, and hence the proceeding itself unfair, even if the errors of counsel cannot be shown by a preponderance of the evidence to have determined the outcome.

Accordingly, the appropriate test for prejudice finds its roots in the test for materiality of exculpatory information not disclosed to the defense by the prosecution, United States v. Agurs, 427 U.S., at 104, 112-113, and in the test for materiality of testimony made unavailable to the defense by Government deportation of a witness, United States v. Valenzuela-Bernal, supra, at 872-874. The defendant must show that there is a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different. A reasonable probability is a probability sufficient to undermine confidence in the outcome.

In making the determination whether the specified errors resulted in the required prejudice, a court should presume, absent challenge to the judgment on grounds of evidentiary insufficiency, that the judge or jury acted according to law. An assessment of the likelihood of a result more favorable to the defendant must exclude the possibility of arbitrariness, whimsy, caprice, "nullification," and the like. A defendant has no entitlement to the luck of a lawless decisionmaker, even if a lawless decision cannot be reviewed. The assessment of prejudice should proceed on the assumption that the decisionmaker is reasonably, conscientiously, and impartially applying the standards that govern the decision. It should not depend on the idiosyncracies of the particular decisionmaker, such as unusual propensities toward harshness or leniency. Although these factors may actually have entered into counsel's selection of strategies and, to that limited extent, may thus affect the performance inquiry, they are irrelevant to the prejudice inquiry. Thus, evidence about the actual process of decision, if not part of the record of the proceeding under review, and evidence about, for example, a particular judge's sentencing practices, should not be considered in the prejudice determination.

The governing legal standard plays a critical role in defining the question to be asked in assessing the prejudice from counsel's errors. When a defendant challenges a conviction, the question is whether there is a reasonable probability that, absent the errors, the factfinder would have had a reasonable doubt respecting guilt. When a defendant challenges a death sentence such as the one at issue in this case, the question is whether there is a reasonable probability that, absent the errors, the sentencer — including an appellate court, to the extent it independently reweighs the evidence — would have concluded that the balance of aggravating and mitigating circumstances did not warrant death.

In making this determination, a court hearing an ineffectiveness claim must consider the totality of the evidence before the judge or jury. Some of the factual findings will have been unaffected by the errors, and factual findings that were affected will have been affected in different ways. Some errors will have had a pervasive effect on the inferences to be drawn from the evidence, altering the entire evidentiary picture, and some will have had an isolated, trivial effect. Moreover, a verdict or conclusion only weakly supported by the record is more likely to have been affected by errors than one with overwhelming record support. Taking the unaffected findings as a given, and taking due account of the effect of the errors on the remaining findings, a court making the prejudice inquiry must ask if the defendant has met the burden of showing that the decision reached would reasonably likely have been different absent the errors.

IV

A number of practical considerations are important for the application of the standards we have outlined. Most important, in adjudicating a claim of actual ineffectiveness of counsel, a court should keep in mind that the principles we have stated do not establish mechanical rules. Although those principles should guide the process of decision, the ultimate focus of inquiry must be on the fundamental fairness of the proceeding whose result is being challenged. In every case the court should be concerned with whether, despite the strong presumption of reliability, the result of the particular proceeding is unreliable because of a breakdown in the adversarial process that our system counts on to produce just results.

To the extent that this has already been the guiding inquiry in the lower courts, the standards articulated today do not require reconsideration of ineffectiveness claims rejected under different standards. Cf. Trapnell v. United States, 725 F.2d, at 153 (in several years of applying "farce and mockery" standard along with "reasonable competence" standard, court "never found that the result of a case hinged on the choice of a particular standard"). In particular, the minor differences in the lower courts' precise formulations of the performance standard are insignificant: the different formulations are mere variations of the overarching reasonableness standard. With regard to the prejudice inquiry, only the strict outcome-determinative test, among the standards articulated in the lower courts, imposes a heavier burden on defendants than the tests laid down today. The difference, however, should alter the merit of an ineffectiveness claim only in the rarest case.

Although we have discussed the performance component of an ineffectiveness claim prior to the prejudice component, there is no reason for a court deciding an ineffective assistance claim to approach the inquiry in the same order or even to address both components of the inquiry if the defendant makes an insufficient showing on one. In particular, a court need not determine whether counsel's performance was deficient before examining the prejudice suffered by the defendant as a result of the alleged deficiencies. The object of an ineffectiveness claim is not to grade counsel's performance. If it is easier to dispose of an ineffectiveness claim on the ground of lack of sufficient prejudice, which we expect will often be so, that course should be followed. Courts should strive to ensure that ineffectiveness claims not become so burdensome to defense counsel that the entire criminal justice system suffers as a result.

The principles governing ineffectiveness claims should apply in federal collateral proceedings as they do on direct appeal or in motions for a new trial. As indicated by the "cause and prejudice" test for overcoming procedural waivers of claims of error, the presumption that a criminal judgment is final is at its strongest in collateral attacks on that judgment. See United States v. Frady, 456 U.S. 152, 162-169 (1982); Engle v. Isaac, 456 U.S. 107, 126-129 (1982). An ineffectiveness claim, however, as our articulation of the standards that govern decision of such claims makes clear, is an attack on the fundamental fairness of the proceeding whose result is challenged. Since fundamental fairness is the central concern of the writ of habeas corpus, see id., at 126, no special standards ought to apply to ineffectiveness claims made in habeas proceedings.

Finally, in a federal habeas challenge to a state criminal judgment, a state court conclusion that counsel rendered effective assistance is not a finding of fact binding on the federal court to the extent stated by 28 U.S.C. § 2254(d). Ineffectiveness is not a question of "basic, primary, or historical fac[t]," Townsend v. Sain, 372 U.S. 293, 309, n. 6 (1963). Rather, like the question whether multiple representation in a particular case gave rise to a conflict of interest, it is a mixed question of law and fact. See Cuyler v. Sullivan, 446 U.S., at 342. Although state court findings of fact made in the course of deciding an ineffectiveness claim are subject to the deference requirement of § 2254(d), and although district court findings are subject to the clearly erroneous standard of Federal Rule of Civil Procedure 52(a), both the performance and prejudice components of the ineffectiveness inquiry are mixed questions of law and fact.

V

Having articulated general standards for judging ineffectiveness claims, we think it useful to apply those standards to the facts of this case in order to illustrate the meaning of the general principles. The record makes it possible to do so. There are no conflicts between the state and federal courts over findings of fact, and the principles we have articulated are sufficiently close to the principles applied both in the Florida courts and in the District Court that it is clear that the factfinding was not affected by erroneous legal principles. See Pullman-Standard v. Swint, 456 U.S. 273, 291-292 (1982).

Application of the governing principles is not difficult in this case. The facts as described above, see supra, at 671-678, make clear that the conduct of respondent's counsel at and before respondent's sentencing proceeding cannot be found unreasonable. They also make clear that, even assuming the challenged conduct of counsel was unreasonable, respondent suffered insufficient prejudice to warrant setting aside his death sentence.

With respect to the performance component, the record shows that respondent's counsel made a strategic choice to argue for the extreme emotional distress mitigating circumstance and to rely as fully as possible on respondent's acceptance of responsibility for his crimes. Although counsel understandably felt hopeless about respondent's prospects, see App. 383-384, 400-401, nothing in the record indicates, as one possible reading of the District Court's opinion suggests, see App. to Pet. for Cert. A282, that counsel's sense of hopelessness distorted his professional judgment. Counsel's strategy choice was well within the range of professionally reasonable judgments, and the decision not to seek more character or psychological evidence than was already in hand was likewise reasonable.

The trial judge's views on the importance of owning up to one's crimes were well known to counsel. The aggravating circumstances were utterly overwhelming. Trial counsel could reasonably surmise from his conversations with respondent that character and psychological evidence would be of little help. Respondent had already been able to mention at the plea colloquy the substance of what there was to know about his financial and emotional troubles. Restricting testimony on respondent's character to what had come in at the plea colloquy ensured that contrary character and psychological evidence and respondent's criminal history, which counsel had successfully moved to exclude, would not come in. On these facts, there can be little question, even without application of the presumption of adequate performance, that trial counsel's defense, though unsuccessful, was the result of reasonable professional judgment.

With respect to the prejudice component, the lack of merit of respondent's claim is even more stark. The evidence that respondent says his trial counsel should have offered at the sentencing hearing would barely have altered the sentencing profile presented to the sentencing judge. As the state courts and District Court found, at most this evidence shows that numerous people who knew respondent thought he was generally a good person and that a psychiatrist and a psychologist believed he was under considerable emotional stress that did not rise to the level of extreme disturbance. Given the overwhelming aggravating factors, there is no reasonable probability that the omitted evidence would have changed the conclusion that the aggravating circumstances outweighed the mitigating circumstances and, hence, the sentence imposed. Indeed, admission of the evidence respondent now offers might even have been harmful to his case: his "rap sheet" would probably have been admitted into evidence, and the psychological reports would have directly contradicted respondent's claim that the mitigating circumstance of extreme emotional disturbance applied to his case.

Our conclusions on both the prejudice and performance components of the ineffectiveness inquiry do not depend on the trial judge's testimony at the District Court hearing. We therefore need not consider the general admissibility of that testimony, although, as noted supra, at 695, that testimony is irrelevant to the prejudice inquiry. Moreover, the prejudice question is resolvable, and hence the ineffectiveness claim can be rejected, without regard to the evidence presented at the District Court hearing. The state courts properly concluded that the ineffectiveness claim was meritless without holding an evidentiary hearing.

Failure to make the required showing of either deficient performance or sufficient prejudice defeats the ineffectiveness claim. Here there is a double failure. More generally, respondent has made no showing that the justice of his sentence was rendered unreliable by a breakdown in the adversary process caused by deficiencies in counsel's assistance. Respondent's sentencing proceeding was not fundamentally unfair.

We conclude, therefore, that the District Court properly declined to issue a writ of habeas corpus. The judgment of the Court of Appeals is accordingly

Reversed.