announced the judgment of the Court and delivered the opinion of the Court with respect to Parts I, II, III-A, IV, and V, and an opinion with respect to Part. III-B, in which The Chief Justice, Justice White, and Justice Scalia join.
The question in this case is whether a court properly may enforce an agreement in which a criminal defendant releases his right to file an action under 42 U. S. C. § 1983 in return for a prosecutor’s dismissal of pending criminal charges.
I.
In 1983, a grand jury m Rockingham County, New Hampshire, indicted David Champy for aggravated felonious sexual assault. Respondent Bernard Rumery, a friend of Champy’s, read about the charges in a local newspaper. Seeking information about the charges, he telephoned Mary Deary, who was acquainted with both Rumery and Champy. Coincidentally, Deary had been the victim of the assault in question and was expected to be the principal witness against Champy. The record does not reveal directly the date or substance of this conversation between Rumery and Deary, but Deary apparently was disturbed by the call. On March 12, according to police records, she called David Barrett, the Chief of Police for the town of Newton. She told him that Rumery was trying to force her to drop the charges against Champy. Rumery talked to Deary again on May 11. The substance of this conversation also is disputed. Rumery claims that Deary called him and that she raised the subject of Champy’s difficulties. According to the police records, however, Deary told Chief Barrett that Rumery had threatened that, if Deary went forward on the Champy case, she would “end up like” two women who recently had been *390murdered in Lowell, Massachusetts. App. 49. Barrett arrested Rumery and accused him of tampering with a witness in violation of N. H. Rev. Stat. Ann. § 641:5(I)(b) (1986), a Class B felony.
Rumery promptly retained Stephen Woods, an experienced criminal defense attorney.1 Woods contacted Brian Graf, the Deputy County Attorney for Rockingham County. He warned Graf that he “had better [dismiss] these charges, because we’re going to win them and after that we’re going to sue.” App. 11. After further discussions, Graf and Woods reached an agreement, under which Graf would dismiss the charges against Rumery if Rumery would agree not to sue the town, its officials, or Deary for any harm caused by the arrest. All parties agreed that one factor in Graf’s decision not to prosecute Rumery was Graf’s desire to protect Deary from the trauma she would suffer if she were forced to testify. As the prosecutor explained in the District Court:
“I had been advised by Chief Barrett that Mary Deary did not want to testify against Mr. Rumery. The witness tampering charge would have required Mary Deary to testify. . . .
“I think that was a particularly sensitive type of case where you are dealing with a victim of an alleged aggravated felonious sexual assault.” Id., at 52 (deposition of Brian Graf).
See also App. to Pet. for Cert. B-2 (District Court’s findings of fact); App. 20 (deposition of defense counsel Woods).
Woods drafted an agreement in which Rumery agreed to release any claims he might have against the town, its officials, or Deary if Graf agreed to dismiss the criminal charges (the release-dismissal agreement). After Graf approved the form of the agreement, Woods presented it to Rumery. Although Rumery’s recollection of the events was quite different, the District Court found that Woods discussed the *391agreement with Rumery in his office for about an hour and explained to Rumery that he would forgo all civil actions if he signed the agreement. Three days later, on June 6, 1983, Rumery returned to Woods’ office and signed the agreement. The criminal charges were dropped.
Ten months later, on April 13, 1984, Rumery filed an action under § 1983 in the Federal District Court for the District of New Hampshire. He alleged that the town and its officers had violated his constitutional rights by arresting him, defaming him, and imprisoning him falsely. The defendants filed a motion to dismiss, relying on the release-dismissal agreement as an affirmative defense. Rumery argued that the agreement was unenforceable because it violated public policy. The court rejected Rumery’s argument and concluded that a “release of claims under section 1983 is valid . . . if it results from a decision that is voluntary, deliberate and informed.” App. to Pet. for Cert. B-6. The court found that Rumery
“is a knowledgeable, industrious individual with vast experience in the business world. . . . [H]e intelligently and carefully, after weighing all the factors, concluded that it would be in his best interest and welfare to sign the covenant. He was also represented by a very competent attorney with more than ordinary expertise in the sometimes complex area of criminal law.” Id., at B-4.
The court then dismissed Rumery’s suit.
On appeal, the Court of Appeals for the First Circuit reversed. It adopted a per se rule invalidating release-dismissal agreements. The court stated:
“It is difficult to envision how release agreements, negotiated in exchange for a decision not to prosecute, serve the public interest. Enforcement of such covenants would tempt prosecutors to trump up charges in reaction to a defendant’s civil rights claim, suppress evidence of police misconduct, and leave unremedied deprivations of constitutional rights.” 778 F. 2d 66, 69 (1985).
*392Because the case raises a question important to the administration of criminal justice, we granted the town’s petition for a writ of certiorari. 475 U. S. 1118 (1986). We reverse.
II
We begin by noting the source of the law that governs this case. The agreement purported to waive a right to sue conferred by a federal statute. The question whether the policies underlying that statute may in some circumstances render that waiver unenforceable is a question of federal law. We resolve this question by reference to traditional common-law principles, as we have resolved other questions about the principles governing §1983 actions. E. g., Pulliam v. Allen, 466 U. S. 522, 539-540 (1984). The relevant principle is well established: a promise is unenforceable if the interest in its enforcement is outweighed in the circumstances by a public policy harmed by enforcement of the agreement.2
III
The Court of Appeals concluded that the public interests related to release-dismissal agreements justified a per se rule of invalidity. We think the court overstated the perceived problems and also failed to credit the significant public interests that such agreements can further. Most importantly, the Court of Appeals did not consider the wide variety of factual situations that can result in release-dismissal agreements. Thus, although we agree that in some cases these agreements may infringe important interests of the criminal defendant and of society as a whole, we do not believe that the mere possibility of harm to these interests calls for a per se rule.
*393A
Rumery’s first objection to release-dismissal agreements is that they are inherently coercive. He argues that it is unfair to present a criminal defendant with a choice between facing criminal charges and waiving his right to sue under § 1983. We agree that some release-dismissal agreements may not be the product of an informed and voluntary decision. The risk, publicity, and expense of a criminal trial may intimidate a defendant, even if he believes his defense is meritorious. But this possibility does not justify invalidating all such agreements. In other contexts criminal defendants are required to make difficult choices that effectively waive constitutional rights. For example, it is well settled that plea bargaining does not violate the Constitution even though a guilty plea waives important constitutional rights. See Brady v. United States, 397 U. S. 742, 752-753 (1970); Santobello v. New York, 404 U. S. 257, 264 (1971) (Douglas, J., concurring).3 We see no reason to believe that release-dismissal agreements pose a more coercive choice than other situations we have accepted. E. g., Corbitt v. New Jersey, 439 U. S. 212 (1978) (upholding a statute that imposed higher sentences on defendants who went to trial than on those who entered guilty pleas). As Justice Harlan explained:
“The criminal process, like the rest of the legal system, is replete with situations requiring ‘the making of difficult judgments’ as to which course to follow. *394McMann v. Richardson, 397 U. S., at 769. Although a defendant may have a right, even of constitutional dimensions, to follow whichever course he chooses, the Constitution does not by that token always forbid requiring him to choose.” Crampton v. Ohio, decided with McGautha v. California, 402 U. S. 183, 213 (1971).
In many cases a defendant’s choice to enter into a release-dismissal agreement will reflect a highly rational judgment that the certain benefits of escaping criminal prosecution exceed the speculative benefits of prevailing in a civil action. Rumery’s voluntary decision to enter this agreement exemplifies such a judgment. Rumery is a sophisticated businessman. He was not in jail and was represented by an experienced criminal lawyer, who drafted the agreement. Rumery considered the agreement for three days before signing it. The benefits of the agreement to Rumery are obvious: he gained immunity from criminal prosecution in consideration of abandoning a civil suit that he may well have lost.
Because Rumery voluntarily waived his right to sue under §1983, the public interest opposing involuntary waiver of constitutional rights is no reason to hold this agreement invalid. Moreover, we find that the possibility of coercion in the making of similar agreements insufficient by itself to justify a per se rule against release-dismissal bargains. If there is such a reason, it must lie in some external public interest necessarily injured by release-dismissal agreements.
B
As we noted above, the Court of Appeals held that all release-dismissal agreements offend public policy because it believed these agreements “tempt prosecutors to trump up charges in reaction to a defendant’s civil rights claim, suppress evidence of police misconduct, and leave unremedied deprivations of constitutional rights.” 778 F. 2d, at 69. We can agree that in some cases there may be a substantial basis for this concern. It is true, of course, that § 1983 actions to *395vindicate civil rights may further significant public interests. But it is important to remember that Rumery had no public duty to institute a § 1983 action merely to further the public’s interest in revealing police misconduct. Congress has confined the decision to bring such actions to the injured individuals, not to the public at large. Thus, we hesitate to elevate more diffused public interests above Rumery’s considered decision that he would benefit personally from the agreement.
We also believe the Court of Appeals misapprehended the range of public interests arguably affected by a release-dismissal agreement. The availability of such agreements may threaten important public interests. They may tempt prosecutors to bring frivolous charges, or to dismiss meritorious charges, to protect the interests of other officials.4 But a per se rule of invalidity fails to credit other relevant public interests and improperly assumes prosecutorial misconduct.5
The vindication of constitutional rights and the exposure of official misconduct are not the only concerns implicated by § 1983 suits. No one suggests that all such suits are meritorious. Many are marginal and some are frivolous. Yet even when the risk of ultimate liability is negligible, the burden of defending such lawsuits is substantial. Counsel may be retained by the official, as well as the governmental entity. Preparation for trial, and the trial itself, will require the time and attention of the defendant officials, to the detriment of *396their public duties. In some cases litigation will extend over a period of years. This diversion of officials from their normal duties and the inevitable expense of defending even unjust claims is distinctly not in the public interest. To the extent release-dismissal agreements protect public officials from the burdens of defending such unjust claims, they further this important public interest.
A per se rule invalidating release-dismissal agreements also assumes that prosecutors will seize the opportunity for wrongdoing. In recent years the Court has considered a number of claims that prosecutors have acted improperly. E. g., Wayte v. United States, 470 U. S. 598 (1985); United States v. Goodwin, 457 U. S. 368 (1982); Bordenkircher v. Hayes, 434 U. S. 357 (1978). Our decisions in those cases uniformly have recognized that courts normally must defer to prosecutorial decisions as to whom to prosecute. The reasons for judicial deference are well known. Prosecutorial charging decisions are rarely simple. In addition to assessing the strength and importance of a case, prosecutors also must consider other tangible and intangible factors, such as government enforcement priorities. See Wayte v. United States, 470 U. S., at 607. Finally, they also must decide how best to allocate the scarce resources of a criminal justice system that simply cannot accommodate the litigation of every serious criminal charge.6 Because these decisions' “are not readily susceptible to the kind of analysis the courts are competent to undertake,” we have been “properly hesitant to examine the decision whether to prosecute.” Id., at 607-608. See United States v. Goodwin, supra, at 373.
*397Against this background of discretion, the mere opportunity to act improperly does not compel an assumption that all — or even a significant number of — release-dismissal agreements stem from prosecutors abandoning “the independence of judgment required by [their] public trust,” Imbler v. Pachtman, 424 U. S. 409, 423 (1976).7 Rather, tradition and experience justify our belief that the great majority of prosecutors will be faithful to their duty. Indeed, the merit of this view is illustrated by this case, where the only evidence of prosecutorial misconduct is the agreement itself.
Because release-dismissal agreements may further legitimate prosecutorial and public interests, we reject the Court of Appeals’ holding that all such agreements are invalid per se.8
IV
Turning to the agreement presented by this case, we conclude that the District Court’s decision to enforce the agreement was correct. As we have noted, supra, at 394, it is *398clear that Rumery voluntarily entered the agreement. Moreover, in this case the prosecutor had an independent, legitimate reason to make this agreement directly related to his prosecutorial responsibilities. The agreement foreclosed both the civil and criminal trials concerning Rumery, in which Deary would have been a key witness. She therefore was spared the public scrutiny and embarrassment she would have endured if she had had to testify in either of those cases.9 Both the prosecutor and the defense attorney testified in the District Court that this was a significant consideration in the prosecutor’s decision. Supra, at 390.
In sum, we conclude that this agreement was voluntary, that there is no evidence of prosecutorial misconduct, and that enforcement of this agreement would not adversely affect the relevant public interests.10
V
We reverse the judgment of the Court of Appeals and remand the case to the District Court for dismissal of the complaint.
It is so ordered.
By the time this case was litigated in the District Court, Woods had become the County Attorney for Rockingham County. App. 51.
Cf. Restatement (Second) of Contracts § 178(1) (1981). See also Crampton v. Ohio, decided with McGautha v. California, 402 U. S. 183, 213 (1971) (“The threshold question is whether compelling [a defendant to decide whether to waive constitutional rights] impairs to an appreciable extent any of the policies behind the rights involved”).
We recognize that the analogy between plea bargains and release-dismissal agreements is not complete. The former are subject to judicial oversight. Moreover, when the State enters a plea bargain with a criminal defendant, it receives immediate and tangible benefits, such as promptly imposed punishment without the expenditure of prosecutorial resources, see Brady v. United States, 397 U. S., at 752. Also, the defendant’s agreement to plead to some crime tends to ensure some satisfaction of the public’s interest in the prosecution of crime and confirms that the prosecutor’s charges have a basis in fact. The benefits the State may realize in particular cases from release-dismissal agreements may not be as tangible, but they are not insignificant.
Actions taken for these reasons properly have been recognized as unethical. See ABA Model Code of Professional Responsibility, Disciplinary Rule 7-105 (1980).
Prosecutors themselves rarely are held liable in §1983 actions. See Imbler v. Pachtman, 424 U. S. 409 (1976) (discussing prosecutorial immunity). Also, in many States and municipalities — perhaps in most — prosecutors are elected officials and are entirely independent of the civil authorities likely to be defendants in § 1983 suits. There may be situations, of course, when a prosecutor is rhotivated to protect the interests of such officials or of police. But the constituency of an elected prosecutor is the public, and such a prosecutor is likely to be influenced primarily by the general public interest.
In 1985, the federal district courts disposed of 47,360 criminal cases. Of these, only 6,053, or about 12.8%, ended after a trial. Annual Report of the Director of the Administrative Office of the U. S. Courts 374 (1985). As we have recognized, if every serious criminal charge were evaluated through a full-scale criminal trial, “the States and the Federal Government would need to multiply by many times the number of judges and court facilities,” Santobello v. New York, 404 U. S. 257, 260 (1971).
Of course, the Court has found that certain actions are so likely to result from prosecutorial misconduct that it has “ ‘presume[d]’ an improper vindictive motive,” United States v. Goodwin, 457 U. S. 368, 373 (1982). E. g., Blackledge v. Perry, 417 U. S. 21 (1974) (holding that it violates the Due Process Clause for a prosecutor to increase charges in response to a defendant’s exercise of his right to appeal). But the complexity of pretrial decisions by prosecutors suggests that judicial evaluation of those decisions should be especially deferential. Thus, the Court has never accepted such a blanket claim with respect to pretrial decisions. See United States v. Goodwin, supra; Bordenkircher v. Hayes, 434 U. S. 357 (1978).
Justice Stevens’ evaluation of the public interests associated with release-dismissal agreements relies heavily on his view that Rumery is a completely innocent man. Post, at 404-407. He rests this conclusion on the testimony Rumery and his attorney presented to the District Court, but fails to acknowledge that the District Court’s factual findings gave little credence to this testimony. Justice Stevens also gives great weight to the fact that Rumery “must be presumed to be innocent.” Post, at 404. But this is not a criminal case. This is a civil case, in which Rumery bears the ultimate burden of proof.
Cf. ABA Standards for Criminal Justice 14 — 1.8(a)(iii) (2d ed. 1980) (following a guilty plea, it is proper for the sentencing judge to consider that the defendant “by making public trial unnecessary, has demonstrated genuine consideration for the victims . . . by . . . preventing] unseemly public scrutiny or embarrassment”).
We note that two Courts of Appeals have applied a voluntariness standard to determine the enforceability of agreements entered into after trial, in which the defendants released possible § 1983 claims in return for sentencing considerations. See Bushnell v. Rossetti, 760 F. 2d 298 (CA4 1984); Jones v. Taber, 648 F. 2d 1201 (CA9 1981). We have no occasion in this case to determine whether an inquiry into voluntariness alone is sufficient to determine the enforceability of release-dismissal agreements. We also note that it would be helpful to conclude release-dismissal agreements under judicial supervision. Although such supervision is not essential to the validity of an otherwise-proper agreement, it would help ensure that the agreements did not result from prosecutorial misconduct.