delivered the opinion of the Court.
We have two questions before us in this case: whether the District Court erred on the merits in rejecting the petition for habeas corpus filed by petitioner, and whether the Court of Appeals for the Fifth Circuit correctly denied a stay of execution of the death penalty pending appeal of the District Court’s judgment.
I
On November 14, 1978, petitioner was convicted of the capital murder of a police officer in Bell County, Tex. A separate sentencing hearing before the same jury was then held to determine whether the death penalty should be imposed. Under Tex. Code Crim. Proc. Ann., Art. 37.071 (Vernon 1981),1 two special questions were to be submitted to the *884jury: whether the conduct causing death was “committed deliberately and with reasonable expectation that the death of the deceased or another would result”; and whether “there is a probability that the defendant would commit criminal acts of violence that would constitute a continuing threat to society.” The State introduced into evidence petitioner’s prior convictions and his reputation for lawlessness. The State also called two psychiatrists, John Holbrook and James Grigson, who, in response to hypothetical questions, testified that petitioner would probably commit further acts of violence and represent a continuing threat to society. The jury answered both of the questions put to them in the affirmative, a result which required the imposition of the death penalty.
On appeal to the Texas Court of Criminal Appeals, petitioner urged, among other submissions, that the use of psychiatrists at the punishment hearing to make predictions *885about petitioner’s future conduct was unconstitutional because psychiatrists, individually and as a class, are not competent to predict future dangerousness. Hence, their predictions are so likely to produce erroneous sentences that their use violated the Eighth and Fourteenth Amendments. It was also urged, in any event, that permitting answers to hypothetical questions by psychiatrists who had not personally examined petitioner was constitutional error. The court rejected all of these contentions and affirmed the conviction and sentence on March 12, 1980, Barefoot v. State, 596 S. W. 2d 875; rehearing was denied on April 30, 1980.
Petitioner’s execution was scheduled for September 17, 1980. On July 29, this Court granted a stay of execution pending the filing and disposition of a petition for certiorari, which was filed and then denied on June 29, 1981. Barefoot v. Texas, 453 U. S. 913. Petitioner’s execution was again scheduled by the state courts, this time for October 13,1981. An application for habeas corpus to the Texas Court of Criminal Appeals was denied on October 7, 1981, whereafter a petition for habeas corpus was filed in the United States District Court for the Western District of Texas. Among other issues, petitioner raised the same claims with respect to the use of psychiatric testimony that he had presented to the state courts. The District Court stayed petitioner’s execution pending action on the petition. An evidentiary hearing was held on July 28, 1982, at which petitioner was represented by competent counsel. On November 9, 1982, the District Court filed its findings and conclusions, rejecting each of the several grounds asserted by petitioner. The writ was accordingly denied; also, the stay of petitioner’s death sentence was vacated. The District Court, however, granted petitioner’s motion to proceed informa pauperis and issued a certificate of probable cause pursuant to 28 U. S. C. § 2253, which provides that an appeal may not be taken to the court of appeals from the final order in a habeas corpus proceeding where the detention complained of arises out of process issued by a state court “unless the justice or judge who *886rendered the order or a circuit justice or judge issues a certificate of probable cause.” Notice of appeal was filed on November 24, 1982.
At this point, the Texas courts set January 25, 1983, as the new execution date. A petition for habeas corpus and motion for stay of execution were then denied by the Texas Court of Criminal Appeals on December 21,1982, and another motion for stay of execution was denied by the same court on January 11, 1983.
On January 14, petitioner moved the Court of Appeals for the Fifth Circuit to stay his execution pending consideration of his appeal from the denial of his petition for habeas corpus. On January 17, the parties were notified to present briefs and oral argument to the court on January 19. The case was heard on January 19, and, on January 20, the Court of Appeals issued an opinion and judgment denying the stay. 697 F. 2d 593 (1983). The court’s opinion recited that the court had studied the briefs and record filed and had heard oral argument at which petitioner’s attorney was allowed unlimited time to discuss any matter germane to the case. The Court of Appeals was of the view that by giving the parties unlimited opportunity to brief and argue the merits as they saw fit, the requirements set forth in this Court’s cases, such as Garrison v. Patterson, 391 U. S. 464 (1968), Nowakowski v. Maroney, 386 U. S. 542 (1967), and Carafas v. LaVallee, 391 U. S. 234 (1968), were satisfied. As the court understood those cases, when a certificate of probable cause is issued by the district court, the court of appeals must give the parties an opportunity to address the merits. In its view, the parties had been given “an unlimited opportunity to make their contentions upon the underlying merits by briefs and oral argument.” 697 F. 2d, at 596. The Court of Appeals then proceeded to address the merits of the psychiatric testimony issue, together with new claims not presented to the District Court, that the state court had no jurisdiction to resentence petitioner and that newly discovered evidence war*887ranted a new trial. Each of the grounds was discussed by the court and rejected. The court concluded that since the petition had no substantial merit, a stay should be denied.
Petitioner then filed an application for stay of execution with the Circuit Justice for the Fifth Circuit, who referred the matter to the Court. On January 24, 1983, the Court stayed petitioner’s execution and, treating the application for stay as a petition for writ of certiorari before judgment, granted certiorari. 459 U. S. 1169. The parties were directed to brief and argue “the question presented by the application, namely, the appropriate standard for granting or denying a stay of execution pending disposition of an appeal by a federal court of appeals by a death-sentenced federal habeas corpus petitioner, and also the issues on appeal before the United States Court of Appeals for the Fifth Circuit.” Ibid. The case was briefed and orally argued here, and we now affirm the judgment of the District Court.
With respect to the procedures followed by the Court of Appeals in refusing to stay petitioner’s death sentence, it must be remembered that direct appeal is the primary avenue for review of a conviction or sentence, and death penalty cases are no exception. When the process of direct review— which, if a federal question is involved, includes the right to petition this Court for a writ of certiorari — comes to an end, a presumption of finality and legality attaches to the conviction and sentence. The role of federal habeas proceedings, while important in assuring that constitutional rights are observed, is secondary and limited. Federal courts are not forums in which to relitigate state trials. Even less is federal habeas a means by which a defendant is entitled to delay an execution indefinitely. The procedures adopted to facilitate the orderly consideration and disposition of habeas petitions are not legal entitlements that a defendant has a right to pursue irrespective of the contribution these procedures make toward *888uncovering constitutional error. “It is natural that counsel for the condemned in a capital case should lay hold of every ground which, in their judgment, might tend to the advantage of their client, but the administration of justice ought not to be interfered with on mere pretexts.” Lambert v. Barrett, 159 U. S. 660, 662 (1895). Furthermore, unlike a term of years, a death sentence cannot begin to be carried out by the State while substantial legal issues remain outstanding. Accordingly, federal courts must isolate the exceptional cases where constitutional error requires retrial or resen-tencing as certainly and swiftly as orderly procedures will permit. They need not, and should not, however, fail to give nonfrivolous claims of constitutional error the careful attention that they deserve.
For these reasons, we granted certiorari before judgment to determine whether the Court of Appeals erred in refusing to stay petitioner’s death sentence.
A
Petitioner urges that the Court of Appeals improperly denied a stay of execution while failing to act finally on his appeal. He suggests the possibility of remanding the case to the Court of Appeals without reaching the merits of the District Court’s judgment. The heart of petitioner’s submission is that the Court of Appeals, unless it believes the case to be entirely frivolous, was obligated to decide the appeal on its merits in the usual course and must, in a death case, stay the execution pending such disposition. The State responds that the Court of Appeals reached and decided the merits of the issues presented in the course of denying the stay and that petitioner had ample opportunity to address the merits.
We have previously held that “if an appellant persuades an appropriate tribunal that probable cause for an appeal exists, he must then be afforded an opportunity to address the underlying merits. ” Garrison v. Patterson, supra, at 466 (per curiam). See Nowakowski v. Maroney, supra; Carafas v. *889LaVallee, supra. These decisions indicate that if a court of appeals is unable to resolve the merits of an appeal before the scheduled date of execution, the petitioner is entitled to a stay of execution to permit due consideration of the merits. But we have also held that the requirement of a decision on the merits “does not prevent the courts of appeals from adopting appropriate summary procedures for final disposition of such cases.” Garrison v. Patterson, 391 U. S., at 466. See Carafas v. LaVallee, 391 U. S., at 242. In Garrison, after examining our prior holdings, we concluded:
“[N]othing [in these cases] prevents the courts of appeals from considering the questions of probable cause and the merits together, and nothing said there or here necessarily requires full briefing in every instance in which a certificate is granted. We hold only that where an appeal possesses sufficient merit to warrant a certificate, the appellant must be afforded adequate opportunity to address the merits, and that if a summary procedure is adopted the appellant must be informed, by rule or otherwise, that his opportunity will be limited.” 391 U. S., at 466.
We emphasized, ibid., that there must be ample evidence that in disposing of the appeal, the merits have been addressed, but that nothing in the cases or the applicable rules prevents a court of appeals from adopting summary procedures in such cases.
On the surface, it is not clear whether the Fifth Circuit’s recent practice of requiring a showing of some prospect of success on the merits before issuing a stay of execution, O’Bryan v. Estelle, 691 F. 2d 706, 708 (1982); Brooks v. Estelle, 697 F. 2d 586 (1982), comports with these requirements. Approving the execution of a defendant before his appeal is decided on the merits would clearly be improper under Garrison, Nowakowski, and Carafas. However, a practice of deciding the merits of an appeal, when possible, *890together with the application for a stay, is not inconsistent with our cases.
It appears clear that the Court of Appeals in this case pursued the latter course. The Court of Appeals was fully aware of our precedents and ruled that their requirements were fully satisfied. After quoting from Garrison, the Court of Appeals said:
“Our actions here fall under this language. Petitioner’s motion is directed solely to the merits. The parties have been also afforded an unlimited opportunity to make their contentions upon the underlying merits and oral argument. This opinion demonstrates the reasons for our decision.” 697 F. 2d, at 596.
In a section of its opinion entitled “Merits of Appeal: Psychiatric Testimony on Dangerousness,” the Court of Appeals then proceeded to address that issue and reject petitioner’s contentions.
The course pursued by the Court of Appeals in this case was within the bounds of our prior decisions. In connection with acting on the stay, the parties were directed to file briefs and to present oral argument. In light of the Fifth Circuit’s announced practice, O’Bryan v. Estelle, supra; Brooks v. Estelle, supra, it was clear that whether a stay would be granted depended on the probability of success on the merits. The parties addressed the merits and were given unlimited time to present argument. We do not agree that petitioner and his attorneys were prejudiced in their preparation of the appeal. The primary issue presented had been briefed and argued throughout the proceedings in the state courts and rebriefed and reargued in the District Court’s habeas corpus proceeding. From the time the District Court ruled on the petition on November 9, 1982, petitioner had 71 days in which to prepare the briefs and arguments which were presented to the Fifth Circuit on January 19, 1983.
*891Although the Court of Appeals did not formally affirm the judgment of the District Court, there is no question that the Court of Appeals ruled on the merits of the appeal, as its concluding statements demonstrate:
“This Court has had the benefit of the full trial court record except for a few exhibits unimportant to our considerations. We have read the arguments and materials filed by the parties. The petitioner is represented here, as he has been throughout the habeas corpus proceedings in state and federal courts, by a competent attorney experienced in this area of the law. We have heard full arguments in open court. Finding no patent substantial merit, or semblance thereof, to petitioner’s constitutional objections, we must conclude and order that the motion for stay should be DENIED.” 697 F. 2d, at 599-600.
It would have been advisable, once the court had addressed the merits and arrived at these conclusions, to verify the obvious by expressly affirming the judgment of the District Court, as well as to deny the stay. The court’s failure to do so, however, does not conflict with Garrison and related cases. Indeed, in Garrison itself, the Court noted that “[i]n an effort to determine whether the merits had been addressed . . . this Court solicited further submissions from the parties in this case.” 391 U. S., at 466, n. 2. If a formal decision on the merits were required, this inquiry would have been pointless. Moreover, the Court of Appeals cannot be faulted for not formally affirming the judgment of the District Court since this Court, over the dissent of three Justices arguing as petitioner does here, refused to stay an execution in a case where the Court of Appeals followed very similar procedures. Brooks v. Estelle, 459 U. S. 1061 (1982).2
*892Although the Court of Appeals moved swiftly to decide the stay, this does not mean that its treatment of the merits was cursory or inadequate. On the contrary, the court’s resolution of the primary issue on appeal, the admission of psychiatric testimony on dangerousness, reflects careful consideration. For these reasons, to remand to the Court of Appeals for verification that the judgment of the District Court was affirmed would be an unwarranted exaltation of form over substance.
B
That the Court of Appeals’ handling of this case was tolerable under our precedents is not to suggest that its course should be accepted as the norm or as the preferred procedure. It is a matter of public record that an increasing number of death-sentenced petitioners are entering the appellate stages of the federal habeas process. The fair and efficient consideration of these appeals requires proper procedures for the handling of applications for stays of executions and demands procedures that allow a decision on the merits of an appeal accompanying the denial of a stay. The development of these procedures is primarily a function of the courts of appeals and the rulemaking processes of the federal courts, but the following general guidelines can be set forth.
First. Congress established the requirement that a prisoner obtain a certificate of probable cause to appeal in order to prevent frivolous appeals from delaying the States’ ability to impose sentences, including death sentences.3 The pri*893mary means of separating meritorious from frivolous appeals should be the decision to grant or withhold a certificate of probable cause. It is generally agreed that “probable cause requires something more than the absence of frivolity and that the standard is a higher one than the ‘good faith’ requirement of § 1915.” Blackmun, Allowance of In Forma Pauperis Appeals in §2255 and Habeas Corpus Cases, 43 F. R. D. 343, 352 (1967). We agree with the weight of opinion in the Courts of Appeals that a certificate of probable cause requires petitioner to make a “substantial showing of the denial of [a] federal right.” Stewart v. Beto, 454 F. 2d 268, 270, n. 2 (CA5 1971), cert. denied, 406 U. S. 925 (1972). See also Ramsey v. Hand, 309 F. 2d 947, 948 (CA10 1962); Goode v. Wainwright, 670 F. 2d 941 (CA11 1982).4 In a capital case, the nature of the penalty is a proper consideration in determining whether to issue a certificate of probable cause, but the severity of the penalty does not in itself suffice to warrant the automatic issuing of a certificate.
Second. When a certificate of probable cause is issued by the district court, as it was in this case, or later by the court of appeals, petitioner must then be afforded an opportunity to address the merits, and the court of appeals is obligated to decide the merits of the appeal. Accordingly, a court of appeals, where necessary to prevent the case from becoming *894moot by the petitioner’s execution, should grant a stay of execution pending disposition of an appeal when a condemned prisoner obtains a certificate of probable cause on his initial habeas appeal.
Third. As our earlier cases have indicated, a court of appeals may adopt expedited procedures in resolving the merits of habeas appeals, notwithstanding the issuance of a certificate of probable cause. If a circuit chooses to follow this course, it would be advisable to promulgate a local rule stating the manner in which such cases will be handled and informing counsel that the merits of an appeal may be decided upon the motion for a stay. Even without special procedures, it is entirely appropriate that an appeal which is “frivolous and entirely without merit” be dismissed after the hearing on a motion for a stay. See, e. g., Local Rule 20, Court of Appeals for the Fifth Circuit. We caution that the issuance of a certificate of probable cause generally should indicate that an appeal is not legally frivolous, and that a court of appeals should be confident that petitioner’s claim is squarely foreclosed by statute, rule, or authoritative court decision, or is lacking any factual basis in the record of the case, before dismissing it as frivolous.
If an appeal is not frivolous, a court of appeals may still choose to expedite briefing and hearing the merits of all or of selected cases in which a stay of a death sentence has been requested, provided that counsel has adequate opportunity to address the merits and knows that he is expected to do so. If appropriate notice is provided, argument on the merits may be heard at the same time the motion for a stay is considered, and the court may thereafter render a single opinion deciding both the merits and the motion, unless exigencies of time preclude a considered decision on the merits, in which case the motion for a stay must be granted. In choosing the procedures to be used, the courts should consider whether the delay that is avoided by summary procedures warrants departing from the normal, untruncated processes of appel*895late review. In instances where expedition of the briefing and argument schedule is not ordered, a court of appeals may nevertheless choose to advance capital cases on the docket so that the decision of these appeals is not delayed by the weight of other business.
Fourth. Second and successive federal habeas corpus petitions present a different issue. “To the extent that these involve the danger that a condemned inmate might attempt to use repeated petitions and appeals as a mere delaying tactic, the State has a quite legitimate interest in preventing such abuses of the writ.” Brief for NAACP Legal Defense and Educational Fund, Inc., as Amicus Curiae 40-41. Title 28 U. S. C. §2254 Rule 9(b) states that “a second or successive petition may be dismissed if the judge finds that it fails to allege new or different grounds for relief ... [or if] the failure of the petitioner to assert those grounds in a prior petition constituted an abuse of the writ.” See Sanders v. United States, 373 U. S. 1, 18 (1963); Advisory Committee Note to Rule 9(b), 28 U. S. C., p. 273. Even where it cannot be concluded that a petition should be dismissed under Rule 9(b), it would be proper for the district court to expedite consideration of the petition. The granting of a stay should reflect the presence of substantial grounds upon which relief might be granted.
Fifth. Stays of execution are not automatic pending the filing and consideration of a petition for a writ of certiorari from this Court to the court of appeals that has denied a writ of habeas corpus. It is well established that there “ 'must be a reasonable probability that four Members of the Court would consider the underlying issue sufficiently meritorious for the grant of certiorari or the notation of probable jurisdiction; there must be a significant possibility of reversal of the lower court’s decision; and there must be a likelihood that irreparable harm will result if that decision is not stayed.’” White v. Florida, 458 U. S. 1301, 1302 (1982) (Powell, J., in chambers) (quoting Times-Picayune Publishing Corp. v. *896Schulingkamp, 419 U. S. 1301, 1305 (1974) (POWELL, J., in chambers)). Applications for stays of death sentences are expected to contain the information and materials necessary to make a careful assessment of the merits of the issue and so reliably to determine whether plenary review and a stay are warranted. A stay of execution should first be sought from the court of appeals, and this Court generally places considerable weight on the decision reached by the courts of appeals in these circumstances.
Ill
Petitioner’s merits submission is that his death sentence must be set aside because the Constitution of the United States barred the testimony of the two psychiatrists who testified against him at the punishment hearing. There are several aspects to this claim. First, it is urged that psychiatrists, individually and as a group, are incompetent to predict with an acceptable degree of reliability that a particular criminal will commit other crimes in the future and so represent a danger to the community. Second, it is said that in any event, psychiatrists should not be permitted to testify about future dangerousness in response to hypothetical questions and without having examined the defendant personally. Third, it is argued that in the particular circumstances of this case, the testimony of the psychiatrists was so unreliable that the sentence should be set aside. As indicated below, we reject each of these arguments.
A
The suggestion that no psychiatrist’s testimony may be presented with respect to a defendant’s future dangerousness is somewhat like asking us to disinvent the wheel. In the first place, it is contrary to our cases. If the likelihood of a defendant’s committing further crimes is a constitutionally acceptable criterion for imposing the death penalty, which it is, Jurek v. Texas, 428 U. S. 262 (1976), and if it is not impossible for even a lay person sensibly to arrive at that conclu*897sion, it makes little sense, if any, to submit that psychiatrists, out of the entire universe of persons who might have an opinion on the issue, would know so little about the subject that they should not be permitted to testify. In Jurek, seven Justices rejected the claim that it was impossible to predict future behavior and that dangerousness was therefore an invalid consideration in imposing the death penalty. Justices Stewart, Powell, and Stevens responded directly to the argument, id., at 274-276:
“It is, of course, not easy to predict future behavior. The fact that such a determination is difficult, however, does not mean that it cannot be made. Indeed, prediction of future criminal conduct is an essential element in many of the decisions rendered throughout our criminal justice system. The decision whether to admit a defendant to bail, for instance, must often turn on a judge’s prediction of the defendant’s future conduct. Any sentencing authority must predict a convicted person’s probable future conduct when it engages in the process of determining what punishment to impose. For those sentenced to prison, these same predictions must be made by parole authorities. The task that a Texas jury must perform in answering the statutory question in issue is thus basically no different from the task performed countless times each day throughout the American system of criminal justice. What is essential is that the jury have before it all possible relevant information about the individual defendant whose fate it must determine. Texas law clearly assures that all such evidence will be adduced.”
Although there was only lay testimony with respect to dangerousness in Jurek, there was no suggestion by the Court that the testimony of doctors would be inadmissable. To the contrary, the joint opinion announcing the judgment said that the jury should be presented with all of the relevant information. Furthermore, in Estelle v. Smith, 451 U. S. 454, 473 *898(1981), in the face of a submission very similar to that presented in this case with respect to psychiatric testimony, we approvingly repeated the above quotation from Jurek and went on to say that we were in “no sense disapproving the use of psychiatric testimony bearing on future dangerousness.” See also California v. Ramos, post, at 1005-1006, 1009-1010, n. 23; Gregg v. Georgia, 428 U. S. 153, 203-204 (1976) (joint opinion) (desirable to allow open and far-ranging argument that places as much information as possible before the jury).
Acceptance of petitioner’s position that expert testimony about future dangerousness is far too unreliable to be admissible would immediately call into question those other contexts in which predictions of future behavior are constantly made. For example, in O’Connor v. Donaldson, 422 U. S. 563, 576 (1975), we held that a nondangerous mental hospital patient could not be held in confinement against his will. Later, speaking about the requirements for civil commitments, we said:
“There may be factual issues in a commitment proceeding, but the factual aspects represent only the beginning of the inquiry. Whether the individual is mentally ill and dangerous to either himself or others and is in need of confined therapy turns on the meaning of the facts which must be interpreted by expert psychiatrists and psychologists.” Addington v. Texas, 441 U. S. 418, 429 (1979).
In the second place, the rules of evidence generally extant at the federal and state levels anticipate that relevant, unprivileged evidence should be admitted and its weight left to the factfinder, who would have the benefit of cross-examination and contrary evidence by the opposing party. Psychiatric testimony predicting dangerousness may be countered not only as erroneous in a particular case but also as generally so unreliable that it should be ignored. If the jury may make up its mind about future dangerousness unaided by psychiatric testimony, jurors should not be barred from hearing the *899views of the State’s psychiatrists along with opposing views of the defendant’s doctors.5
Third, petitioner’s view mirrors the position expressed in the amicus brief of the American Psychiatric Association (APA). As indicated above, however, the same view was presented and rejected in Estelle v. Smith. We are no more convinced now that the view of the APA should be converted into a constitutional rule barring an entire category of expert testimony.6 We are not persuaded that such testimony is almost entirely unreliable and that the factfinder and the adversary system will not be competent to uncover, recognize, and take due account of its shortcomings.
The amicus does not suggest that there are not other views held by members of the Association or of the profession generally. Indeed, as this case and others indicate, there are those doctors who are quite willing to testify at the sentencing hearing, who think, and will say, that they know what they are talking about, and who expressly disagree with the Association’s point of view.7 Furthermore, their *900qualifications as experts are regularly accepted by the courts. If they are so obviously wrong and should be discredited, there should be no insuperable problem in doing so by calling *901members of the Association who are of that view and who confidently assert that opinion in their amicus brief. Neither petitioner nor the Association suggests that psychiatrists are always wrong with respect to future dangerousness, only most of the time. Yet the submission is that this category of testimony should be excised entirely from all trials. We are unconvinced, however, at least as of now, that the adversary process cannot be trusted to sort out the reliable from the unreliable evidence and opinion about future dangerousness, particularly when the convicted felon has the opportunity to present his own side of the case.
We are unaware of and have not been cited to any case, federal or state, that has adopted the categorical views of the Association.8 Certainly it was presented and rejected at every *902stage of the present proceeding. After listening to the two schools of thought testify not only generally but also about the petitioner and his criminal record, the District Court found:
“The majority of psychiatric experts agree that where there is a pattern of repetitive assaultive and violent conduct, the accuracy of psychiatric predictions of future dangerousness dramatically rises. The accuracy of this conclusion is reaffirmed by the expert medical testimony in this case at the evidentiary hearing. ... It would appear that Petitioner’s complaint is not the diagnosis and prediction made by Drs. Holbrook and Grigson at the punishment phase of his trial, but that Dr. Grigson expressed extreme certainty in his diagnosis and prediction. ... In any event, the differences among the experts were quantitative, not qualitative. The differences in opinion go to the weight [of the evidence] and not the admissibility of such testimony. . . . Such disputes are within the province of the jury to resolve. Indeed, it is a fundamental premise of our entire system of criminal jurisprudence that the purpose of the jury is to sort out the true testimony from the false, the important matters from the unimportant matters, and, when called upon to do so, to give greater credence to one party’s expert witnesses than another’s. Such matters occur routinely in the American judicial system, both civil and criminal.” App. 13-14 (footnote omitted).
*903We agree with the District Court, as well as with the Court of Appeals’ judges who dealt with the merits of the issue and agreed with the District Court in this respect.
B
Whatever the decision may be about the use of psychiatric testimony, in general, on the issue of future dangerousness, petitioner urges that such testimony must be based on personal examination of the defendant and may not be given in response to hypothetical questions. We disagree. Expert testimony, whether in the form of an opinion based on hypothetical questions or otherwise, is commonly admitted as evidence where it might help the factfinder do its assigned job. As the Court said long ago in Spring Co. v. Edgar, 99 U. S. 645, 657 (1879):
“Men who have made questions of skill or science the object of their particular study, says Phillips, are competent to give their opinions in evidence. Such opinions ought, in general, to be deduced from facts that are not disputed, or from facts given in evidence; but the author proceeds to say that they need not be founded upon their own personal knowledge of such facts, but may be founded upon the statement of facts proved in the case. Medical men, for example, may give their opinions not only as to the state of a patient they may have visited, or as to the cause of the death of a person whose body they have examined, or as to the nature of the instruments which caused the wounds they have examined, but also in cases where they have not themselves seen the patient, and have only heard the symptoms and particulars of his state detailed by other witnesses at the trial. Judicial tribunals have in many instances held that medical works are not admissible, but they everywhere hold that men skilled in science, art, or particular trades may give their opinions as witnesses in matters pertaining to their professional calling.”
*904See also Dexter v. Hall, 15 Wall. 9, 26-27 (1873); Forsyth v. Doolittle, 120 U. S. 73, 78 (1887); Bram v. United States, 168 U. S. 532, 568-569 (1897).
Today, in the federal system, Federal Rules of Evidence 702-706 provide for the testimony of experts. The Advisory Committee Notes touch on the particular objections to hypothetical questions, but none of these caveats lends any support to petitioner’s constitutional arguments. Furthermore, the Texas Court of Criminal Appeals could find no fault with the mode of examining the two psychiatrists under Texas law:
“The trial court did not err by permitting the doctors to testify on the basis of the hypothetical question. The use of hypothetical questions in the examination of expert witnesses is a well-established practice. 2 C. McCormick and R. Ray, Texas Evidence, § 1402 (2d ed. 1956). That the experts had not examined appellant went to the weight of their testimony, not to its admissibility.” 596 S. W. 2d, at 887.
Like the Court of Criminal Appeals, the District Court, and the Court of Appeals, we reject petitioner’s constitutional arguments against the use of hypothetical questions. Although cases such as this involve the death penalty, we perceive no constitutional barrier to applying the ordinary rules of evidence governing the use of expert testimony.
C
As we understand petitioner, he contends that even if the use of hypothetical questions in predicting future dangerousness is acceptable as a general rule, the use made of them in his case violated his right to due process of law. For example, petitioner insists that the doctors should not have been permitted to give an opinion on the ultimate issue before the jury, particularly when the hypothetical questions
*905were phrased in terms of petitioner’s own conduct;9 that the hypothetical questions referred to controverted facts;10 and that the answers to the questions were so positive as to be assertions of fact and not opinion.11 These claims of misuse of the hypothetical questions, as well as others, were rejected by the Texas courts, and neither the District Court nor the Court of Appeals found any constitutional infirmity in the application of the Texas Rules of Evidence in this particular case. We agree.
t> i — j
In sum, we affirm the judgment of the District Court. There is no doubt that the psychiatric testimony increased the likelihood that petitioner would be sentenced to death, but this fact does not make that evidence inadmissible, any more than it would with respect to other relevant evidence *906against any defendant in a criminal case. At bottom, to agree with petitioner’s basic position would seriously undermine and in effect overrule Jurek v. Texas, 428 U. S. 262 (1976). Petitioner conceded as much at oral argument. Tr. of Oral Arg. 23-25. We are not inclined, however, to overturn the decision in that case.
The judgment of the District Court is
Affirmed.
Texas Code Crim. Proc. Ann., Art. 37.071 (Vernon 1981), provides:
“(a) Upon a finding that the defendant is guilty of a capital offense, the court shall conduct a separate sentencing proceeding to determine whether *884the defendant shall be sentenced to death or life imprisonment. The proceeding shall be conducted in the trial court before the trial jury as soon as practicable. In the proceeding, evidence may be presented as to any matter that the court deems relevant to sentence. This subsection shall not be construed to authorize the introduction of any evidence secured in violation of the Constitution of the United States or of the State of Texas. The state and the defendant or his counsel shall be permitted to present argument for or against sentence of death.
“(b) On conclusion of the presentation of the evidence, the court shall submit the following issues to the jury:
“(1) whether the conduct of the defendant that caused the death of the deceased was committed deliberately and with the reasonable expectation that the death of the deceased or another would result;
“(2) whether there is a probability that the defendant would commit criminal acts of violence that would constitute a continuing threat to society; and “(3) if raised by the evidence, whether the conduct of the defendant in killing the deceased was unreasonable in response to the provocation, if any, by the deceased.
“(c) The state must prove each issue submitted beyond a reasonable doubt, and the jury shall return a special verdict of ‘yes’ or ‘no’ on each issue submitted.”
The question specified in (b)(3) was not submitted to the jury.
In that case, we treated the application for stay as a petition for certio-rari or in the alternative as a petition for certiorari before judgment. We denied the petition on either assumption.
The Habeas Corpus Act of 1867, Act of Feb. 5,1867, ch. 28, § 1,14 Stat. 385, the first Act empowering federal courts to issue a writ of habeas corpus for persons in state custody, imposed an automatic stay of “any proceeding against such person” pending “such proceedings or appeal” involved in determination of a prisoner’s petition. Id., at 386; see Rev. Stat. § 766. This provision required a stay of execution pending disposition of an appeal in capital cases. Rogers v. Peck, 199 U. S. 425, 436 (1905); Lambert v. Barrett, 159 U. S. 660, 662 (1895). In 1908, concerned with the increasing number of frivolous habeas corpus petitions challenging capital sentences which delayed execution pending completion of the appellate process, Con*893gress inserted the requirement that a prisoner first obtain a certificate of probable cause to appeal before being entitled to do so. Act of Mar. 10, 1908, ch. 76, 35 Stat. 40. See H. R. Rep. No. 23, 60th Cong., 1st Sess., 1-2 (1908); 42 Cong. Rec. 608-609 (1908).
The following quotation cogently sums up this standard:
“In requiring a ‘question of some substance’, or a ‘substantial showing of the denial of [a] federal right,’ obviously the petitioner need not show that he should prevail on the merits. He has already failed in that endeavor. Rather, he must demonstrate that the issues are debatable among jurists of reason; that a court could resolve the issues [in a different manner]; or that the questions are ‘adequate to deserve encouragement to proceed further.’” Gordon v. Willis, 516 F. Supp. 911, 913 (ND Ga. 1980) (citing United States ex rel. Jones v. Richmond, 245 F. 2d 234 (CA2), cert. denied, 355 U. S. 846 (1957)).
In this ease, no evidence was offered by petitioner at trial to contradict the testimony of Doctors Holbrook and Grigson. Nor is there a contention that, despite petitioner’s claim of indigence, the court refused to provide an expert for petitioner. In cases of indigency, Texas law provides for the payment of $500 for “expenses incurred for purposes of investigation and expert testimony.” Tex. Code Crim. Proe. Ann., Art. 26.05(d) (Vernon Supp. 1982).
The federal cases cited in Justice Blackmun’s dissent as rejecting “scientific proof,” post, at 931, n. 9, are not constitutional decisions, but decisions of federal evidence law. The question before us is whether the Constitution forbids exposing the jury or judge in a state criminal trial to the opinions of psychiatrists about an issue that Justice Blackmun’s dissent concedes the factfinders themselves are constitutionally competent to decide.
At trial, Dr. Holbrook testified without contradiction that a psychiatrist could predict the future dangerousness of an individual, if given enough background information about the individual. Tr. of Trial (T. Tr.) 2072-2073. Dr. Grigson obviously held a similar view. See id., at 2110, 2134. At the District Court hearing on the habeas petition, the State called two expert witnesses, Dr. George Parker, a psychologist, and Dr. Richard *900Koons, a psychiatrist. Both of these doctors agreed that accurate predictions of future dangerousness can be made if enough information is provided; furthermore, they both deemed it highly likely that an individual fitting the characteristics of the one in the Barefoot hypothetical would commit future acts of violence. Tr. of Hearing (H. Tr.) 183-248.
Although Barefoot did not present any expert testimony at his trial, at the habeas hearing he called Dr. Fred Fason, a psychiatrist, and Dr. Wendell Dickerson, a psychologist. Dr. Fason did not dwell on the general ability of mental health professionals to predict future dangerousness. Instead, for the most part, he merely criticized the giving of a diagnosis based upon a hypothetical question, without an actual examination. He conceded that, if a medical student described a patient in the terms of the Barefoot hypothetical, his “highest order of suspicion,” to the degree of 90%, would be that the patient had a sociopathic personality. Id., at 22. He insisted, however, that this was only an “initial impression,” and that no doctor should give a firm “diagnosis” without a full examination and testing. Id., at 22, 29-30, 36. Dr. Dickerson, petitioner’s other expert, was the only person to testify who suggested that no reliable psychiatric predictions of dangerousness could ever be made.
We are aware that many mental health professionals have questioned the usefulness of psychiatric predictions of future dangerousness in light of studies indicating that such predictions are often inaccurate. For example, at the habeas hearing, Dr. Dickerson, one of petitioner’s expert witnesses, testified that psychiatric predictions of future dangerousness were wrong two out of three times. Id., at 97, 108. He conceded, however, that, despite the high error rate, one “excellently done” study had shown “some predictive validity for predicting violence.” Id., at 96, 97. Dr. John Monahan, upon whom one of the State’s experts relied as “the leading thinker on this issue,” id., at 195, concluded that “the ‘best’ clinical research currently in existence indicates that psychiatrists and psychologists are accurate in no more than one out of three predictions of violent behavior over a several-year period among institutionalized populations that had both committed violence in the past. . . and who were diagnosed as mentally ill.” J. Monahan, The Clinical Prediction of Violent Behavior 47-49 (1981) (emphasis in original). However, although Dr. Monahan originally believed that it was impossible to predict violent behavior, by the time he had completed his monograph, he felt that “there may be circumstances in which prediction is both empirically possible and ethically *901appropriate,” and he hoped that his work would improve the appropriateness and accuracy of clinical predictions. Id., at v.
All of these professional doubts about the usefulness of psychiatric predictions can be called to the attention of the jury. Petitioner’s entire argument, as well as that of Justice Blackmun’s dissent, is founded on the premise that a jury will not be able to separate the wheat from the chaff. We do not share in this low evaluation of the adversary process.
Petitioner relies on People v. Murtishaw, 29 Cal. 3d 733, 631 P. 2d 446 (1981). There the California Supreme Court held that in light of the general unreliability of such testimony, admitting medical testimony concerning future dangerousness was error in the context of a sentencing proceeding under the California capital punishment statutes. The court observed that “the testimony of [the psychiatrist was] not relevant to any of the listed factors” which the jury was to consider in deciding whether to impose the death penalty. Id., at 771-772, 631 P. 2d, at 469. The court distinguished eases, however, where “the trier of fact is required by statute to determine whether a person is ‘dangerous,’ ” in which event “expert prediction, unreliable though it may be, is often the only evidence available to assist the trier of fact. ” Ibid. Furthermore, the court acknowledged that “despite the recognized general unreliability of predictions concerning future violence, it may be possible for a party in a particular case to show that a reliable prediction is possible. ... A reliable prediction might also be conceivable if the defendant had exhibited a long-continued pattern of criminal violence such that any knowledgeable psychiatrist would anticipate future violence.” Id., at 774, 631 P. 2d, at 470. Finally, we note *902that the court did not in any way indicate that its holding was based on constitutional grounds.
Petitioner also relies on White v. Estelle, 564 P. Supp. 851 (SD Tex.1982). The court in that case did no more than express “serious reservations” about the use of psychiatric predictions based on hypotheticals in instances where the doctor has had no previous contact with the defendant. Id., at 858. The actual holding of the case, which is totally irrelevant to the issues here, was that the testimony of a doctor who had interviewed the defendant should have been excluded because, prior to the interview, the defendant had not been given Miranda warnings or an opportunity to consult with his attorney, as required by Estelle v. Smith, 451 U. S. 454 (1981).
There is support for this view in our cases, United States v. Spaulding, 293 U. S. 498, 506 (1935), but it does not appear from what the Court there said that the rule was rooted in the Constitution. In any event, we note that the Advisory Committee Notes to Rule 704 of the Federal Rules of Evidence state as follows:
“The basic approach to opinions, lay and expert, in these rules is to admit them when helpful to the trier of fact. In order to render this approach fully effective and to allay any doubt on the subject, the so-called ‘ultimate issue’ rule is abolished by the instant rule.” 28 U. S. C. App., p. 571.
Nothing prevented petitioner from propounding a hypothetical to the doctors based on his own version of the facts. On cross-examination, both Drs. Holbrook and Grigson readily admitted that their opinions might change if some of the assumptions in the State’s hypothetical were not true. T. Tr. 2104, 2132-2133.
The more certain a State’s expert is about his prediction, the easier it is for the defendant to impeach him. For example, in response to Dr. Grigson’s assertion that he was “100% sure” that an individual with the characteristics of the one in the hypothetical would commit acts of violence in the future, Dr. Fason testified at the habeas hearing that if a doctor claimed to be 100% sure of something without examining the patient, “we would kick him off the staff of the hospital for his arrogance.” H. Tr. 48. Similar testimony could have been presented at Barefoot’s trial, but was not.