delivered the opinion of the Court.
In this case, we consider the showing necessary for a defendant to be entitled to discovery on a claim that the prosecuting attorney singled him out for prosecution on the basis of his race. We conclude that respondents failed to satisfy the threshold showing: They failed to show that the Government declined to prosecute similarly situated suspects of other races.
In April 1992, respondents were indicted in the United States District Court for the Central District of California on charges of conspiring to possess with intent to distribute more than 50 grams of cocaine base (crack) and conspiring to distribute the same, in violation of 21 U. S. C. §§841 and 846 (1988 ed. and Supp. IV), and federal firearms offenses. For three months prior to the indictment, agents of the Federal Bureau of Alcohol, Tobacco, and Firearms and the Narcotics Division of the Inglewood, California, Police Department had infiltrated a suspected crack distribution ring by using three confidential informants. On seven separate occasions during this period, the informants had bought a total of 124.3 grams of crack from respondents and witnessed respondents carrying firearms during the sales. ■ The agents searched the hotel room in which the sales were transacted, arrested respondents Armstrong and Hampton in the room, and found *459more crack and a loaded gun. The agents later arrested the other respondents as part of the ring.
In response to the indictment, respondents filed a motion for discovery or for dismissal of the indictment, alleging that they were selected for federal prosecution because they are black. In support of their motion, they offered only an affidavit by a “Paralegal Specialist,” employed by the Office of the Federal Public Defender representing one of the respondents. The only allegation in the affidavit was that, in every one of the 24 § 841 or § 846 cases closed by the office during 1991, the defendant was black. Accompanying the affidavit was a “study” listing the 24 defendants, their race, whether they were prosecuted for dealing cocaine as well as crack, and the status of each case.1
The Government opposed the discovery motion, arguing, among other things, that there was no evidence or allegation “that the Government has acted unfairly or has prosecuted non-black defendants or failed to prosecute them.” App. 150. The District Court granted the motion. It ordered the Government (1) to provide a list of all cases from the last three years in which the Government charged both cocaine and firearms offenses, (2) to identify the race of the defendants in those cases, (3) to identify what levels of law enforcement were involved in the investigations of those cases, and (4) to explain its criteria for deciding to prosecute those defendants for federal cocaine offenses. Id., at 161-162.
The Government moved for reconsideration of the District Court’s discovery order. With this motion it submitted af*460fidavits and other evidence to explain why it had chosen to prosecute respondents and why respondents’ study did not support the inference that the Government was singling out blacks for cocaine prosecution. The federal and local agents participating in the case alleged in affidavits that race played no role in their investigation. An Assistant United States Attorney explained in an affidavit that the decision to prosecute met the general criteria for prosecution, because
“there was over 100 grams of cocaine base involved, over twice the threshold necessary for a ten year mandatory minimum sentence; there were multiple sales involving multiple defendants, thereby indicating a fairly substantial crack cocaine ring; . . . there were multiple federal firearms violations intertwined with the narcotics trafficking; the overall evidence in the case was extremely strong, including audio and videotapes of defendants; ... and several of the defendants had criminal histories including narcotics and firearms violations.” Id., at 81.
The Government also submitted sections of a published 1989 Drug Enforcement Administration report which concluded that “[l]arge-scale, interstate trafficking networks controlled by Jamaicans, Haitians and Black street gangs dominate the manufacture and distribution of crack.” J. Featherly & E. Hill, Crack Cocaine Overview 1989; App. 103.
In response, one of respondents’ attorneys submitted an affidavit alleging that an intake coordinator at a drug treatment center had told her that there are “an equal number of Caucasian users and dealers to minority users and dealers.” Id., at 138. Respondents also submitted an affidavit from a criminal defense attorney alleging that in his experience many nonblacks are prosecuted in state court for crack offenses, id., at 141, and a newspaper article reporting that federal “crack criminals . . . are being punished far more severely than if they had been caught with powder cocaine, *461and almost every single one of them is black,” Newton, Harsher Crack Sentences Criticized as Racial Inequity, Los Angeles Times, Nov. 23, 1992, p. 1; App. 208-210.
The District Court denied the motion for reconsideration. When the Government indicated it would not comply with the court’s discovery order, the court dismissed the case.2
A divided three-judge panel of the Court of Appeals for the Ninth Circuit reversed, holding that, because of the proof requirements for a selective-prosecution claim, defendants must “provide a colorable basis for believing that ‘others similarly situated have not been prosecuted’ ” to obtain discovery. 21 F. 3d 1431, 1436 (1994) (quoting United States v. Wayte, 710 F. 2d 1385, 1387 (CA9 1983), aff’d, 470 U. S. 598 (1985)). The Court of Appeals voted to rehear the case en banc, and the en banc panel affirmed the District Court’s order of dismissal, holding that “a defendant is not required to demonstrate that the government has failed to prosecute others who are similarly situated.” 48 F. 3d 1508, 1516 (1995) (emphasis deleted). We granted certiorari to determine the appropriate standard for discovery for a selective-prosecution claim. 516 U. S. 942 (1995).
Neither the District Court nor the Court of Appeals mentioned Federal Rule of Criminal Procedure 16, which by its terms governs discovery in criminal cases. Both parties now discuss the Rule in their briefs, and respondents contend that it supports the result reached by the Court of Appeals. Rule 16 provides, in pertinent part:
“Upon request of the defendant the government shall permit the defendant to inspect and copy or photograph books, papers, documents, photographs, tangible objects, *462buildings or places, or copies or portions thereof, which are within the possession, custody or control of the government, and which are material to the preparation of the defendant’s defense or are intended for use by the government as evidence in chief at the trial, or were obtained from or belong to the defendant.” Fed. Rule Grim. Proc. 16(a)(1)(C).
Respondents argue that documents “within the possession ... of the government” that discuss the Government’s prosecution strategy for cocaine cases are “material” to respondents’ selective-prosecution claim. Respondents argue that the Rule applies because any claim that “results in noncon-viction” if successful is a “defense” for the Rule’s purposes, and a successful selective-prosecution claim has that effect. Tr. of Oral Arg. 30.
We reject this argument, because we conclude that in the context of Rule 16 “the defendant’s defense” means the defendant’s response to the Government’s case in chief. While it might be argued that as a general matter, the concept of a “defense” includes any claim that is a “sword,” challenging the prosecution’s conduct of the case, the term may encompass only the narrower class of “shield” claims, which refute the Government’s arguments that the defendant committed the crime charged. Rule 16(a)(1)(C) tends to support the “shield-only” reading. If “defense” means an argument in response to the prosecution’s case in chief, there is a perceptible symmetry between documents “material to the preparation of the defendant’s defense,” and, in the very next phrase, documents “intended for use by the government as evidence in chief at the trial.”
If this symmetry were not persuasive enough, subdivision (a)(2) of Rule 16 establishes beyond peradventure that “defense” in subdivision (a)(1)(C) can refer only to defenses in response to the Government’s case in chief. Rule 16(a)(2), as relevant here, exempts from defense inspection “reports, memoranda, or other internal government documents made *463by the attorney for the government or other government agents in connection with the investigation or prosecution of the case.”
Under Rule 16(a)(1)(C), a defendant may examine documents material to his defense, but, under Rule 16(a)(2), he may not examine Government work product in connection with his case. If a selective-prosecution claim is a “defense,” Rule 16(a)(1)(C) gives the defendant the right to examine Government work product in every prosecution except his own. Because respondents’ construction of “defense” creates the anomaly of a defendant’s being able to examine all Government work product except the most pertinent, we find their construction implausible. We hold that Rule 16(a)(1)(C) authorizes defendants to examine Government documents material to the preparation of their defense against the Government’s case in chief, but not to the preparation of selective-prosecution claims.
In Wade v. United States, 504 U. S. 181 (1992), we considered whether a federal court may review a Government decision not to file a motion to reduce a defendant’s sentence for substantial assistance to the prosecution, to determine whether the Government based its decision on the defendant’s race or religion. In holding that such a decision was reviewable, we assumed that discovery would be available if the defendant could make the appropriate threshold showing, although we concluded that the defendant in that case did not make such a showing. See id., at 186. We proceed on a like assumption here.
A selective-prosecution claim is not a defense on the merits to the criminal charge itself, but an independent assertion that the prosecutor has brought the charge for reasons forbidden by the Constitution. Our cases delineating the necessary elements to prove a claim of selective prosecution have taken great pains to explain that the standard is a demanding one. These cases afford a “background presumption,” cf. United States v. Mezzanatto, 513 U. S. 196, 203 *464(1995), that the showing necessary to obtain discovery should itself be a significant barrier to the litigation of insubstantial claims.
A selective-prosecution claim asks a court to exercise judicial power over a “special province” of the Executive. Heckler v. Chaney, 470 U. S. 821, 832 (1985). The Attorney General and United States Attorneys retain “ ‘broad discretion’ ” to enforce the Nation’s criminal laws. Wayte v. United States, 470 U. S. 598, 607 (1985) (quoting United States v. Goodwin, 457 U. S. 368, 380, n. 11 (1982)). They have this latitude because they are designated by statute as the President’s delegates to help him discharge his constitutional responsibility to “take Care that the Laws be faithfully executed.” U. S. Const., Art. II, § 3; see 28 U. S. C. §§ 516, 547. As a result, “[t]he presumption of regularity supports” their prosecutorial decisions and, “in the absence of clear evidence to the contrary, courts presume that they have properly discharged their official duties.” United States v. Chemical Foundation, Inc., 272 U. S. 1, 14-15 (1926). In the ordinary case, “so long as the prosecutor has probable cause to believe that the accused committed an offense defined by statute, the decision whether or not to prosecute, and what charge to file or bring before a grand jury, generally rests entirely in his discretion.” Bordenkircher v. Hayes, 434 U. S. 357, 364 (1978).
Of course, a prosecutor’s discretion is “subject to constitutional constraints.” United States v. Batchelder, 442 U. S. 114, 125 (1979). One of these constraints, imposed by the equal protection component of the Due Process Clause of the Fifth Amendment, Bolling v. Sharpe, 347 U. S. 497, 500 (1954), is that the decision whether to prosecute may not be based on “an unjustifiable standard' such as race, religion, or other arbitrary classification,” Oyler v. Boles, 368 U. S. 448, 456 (1962). A defendant may demonstrate that the administration of a criminal law is “directed so exclusively against a particular class of persons . . . with a mind so unequal and *465oppressive” that the system of prosecution amounts to “a practical denial” of equal protection of the law. Yick Wo v. Hopkins, 118 U. S. 356, 373 (1886).
In order to dispel the presumption that a prosecutor has not violated equal protection, a criminal defendant must present “clear evidence to the contrary.” Chemical Foundation, supra, at 14-15. We explained in Wayte why courts are “properly hesitant to examine the decision whether to prosecute.” 470 U. S., at 608. Judicial deference to the decisions of these executive officers rests in part on an assessment of the relative competence of prosecutors and courts. “Such factors as the strength of the case, the prosecution’s general deterrence value, the Government’s enforcement priorities, and the case’s relationship to the Government’s overall enforcement plan are not readily susceptible to the kind of analysis the courts are competent to undertake.” Id., at 607. It also stems from a concern not to unnecessarily impair the performance of a core executive constitutional function. “Examining the basis of a prosecution delays the criminal proceeding, threatens to chill law enforcement by subjecting the prosecutor’s motives and decisionmaking to outside inquiry, and may undermine prosecutorial effectiveness by revealing the Government’s enforcement policy.” Ibid.
The requirements for a selective-prosecution claim draw on “ordinary equal protection standards.” Id., at 608. The claimant must demonstrate that the federal prosecutorial policy “had a discriminatory effect and that it was motivated by a discriminatory purpose.” Ibid.; accord, Oyler, supra, at 456. To establish a discriminatory effect in a race case, the claimant must show that similarly situated individuals of a different race were not prosecuted. This requirement has been established in our case law since Ah Sin v. Wittman, 198 U. S. 500 (1905). Ah Sin, a subject of China, petitioned a California state court for a writ of habeas corpus, seeking discharge from imprisonment under a San Francisco County *466ordinance prohibiting persons from setting up gambling tables in rooms barricaded to stop police from entering. Id., at 503. He alleged in his habeas petition “that the ordinance is enforced ‘solely and exclusively against persons of the Chinese race and not otherwise.’ ” Id., at 507. We rejected his contention that this averment made out a claim under the Equal Protection Clause, because it did not allege “that the conditions and practices to which the ordinance was directed did not exist exclusively among the Chinese, or that there were other offenders against the ordinance than the Chinese as to whom it was not enforced.” Id., at 507-508.
The similarly situated requirement does not make a selective-prosecution claim impossible to prove. Twenty years before Ah Sin, we invalidated an ordinance, also adopted by San Francisco, that prohibited the operation of laundries in wooden buildings. Yick Wo, 118 U. S., at 374. The plaintiff in error successfully demonstrated that the ordinance was applied against Chinese nationals but not against other laundry-shop operators. The authorities had denied the applications of 200 Chinese subjects for permits to operate shops in wooden buildings, but granted the applications of 80 individuals who were not Chinese subjects to operate laundries in wooden buildings “under similar conditions.” Ibid. We explained in Ah Sin why the similarly situated requirement is necessary:
“No latitude of intention should be indulged in a case like this. There should be certainty to every intent. Plaintiff in error seeks to set aside a criminal law of the State, not on the ground that it is unconstitutional on its face, not that it is discriminatory in tendency and ultimate actual operation as the ordinance was which was passed on in the Yick Wo case, but that it was made so by the manner of its administration. This is a matter of proof, and no fact should be omitted to make it out completely, when the power of a Federal court is in*467voked to interfere with the course of criminal justice of a State.” 198 U. S., at 508 (emphasis added).
Although Ah Sin involved federal review of a state conviction, we think a similar rule applies where the power of a federal court is invoked to challenge an exercise of one of the core powers of the Executive Branch of the Federal Government, the power to prosecute.
Respondents urge that cases such as Batson v. Kentucky, 476 U. S. 79 (1986), and Hunter v. Underwood, 471 U. S. 222 (1985), cut against any absolute requirement that there be a showing of failure to prosecute similarly situated individuals. We disagree. In Hunter, we invalidated a state law disenfranchising persons convicted of crimes involving moral turpitude. Id., at 233. Our holding was consistent with ordinary equal protection principles, including the similarly situated requirement. There was convincing direct evidence that the State had enacted the provision for the purpose of disenfranchising blacks, id., at 229-231, and indisputable evidence that the state law had a discriminatory effect on blacks as compared to similarly situated whites: Blacks were “ ‘by even the most modest estimates at least 1.7 times as likely as whites to suffer disfranchisement under’” the law in question, id., at 227 (quoting Underwood v. Hunter, 730 F. 2d 614, 620 (CA11 1984)). Hunter thus affords no support for respondents’ position.
In Batson, we considered “[t]he standards for assessing a prima facie case in the context of discriminatory selection of the venire” in a criminal trial. 476 U. S., at 96. We required a criminal defendant to show “that the prosecutor has exercised peremptory challenges to remove from the venire members of the defendant’s race” and that this fact, the potential for abuse inherent in a peremptory strike, and “any other relevant circumstances raise an inference that the prosecutor used that practice to exclude the veniremen from the petit jury on account of their race.” Ibid. During jury selection, the entire res gestae take place in front of the trial *468judge. Because the judge has before him the entire venire, he is well situated to detect whether a challenge to the seating of one juror is part of a “pattern” of singling out members of a single race for peremptory challenges. See id., at 97. He is in a position to discern whether a challenge to a black juror has evidentiary significance; the significance may differ if the venire consists mostly of blacks or of whites. Similarly, if the defendant makes out a prima facie case, the prosecutor is called upon to justify only decisions made in the very case then before the court. See id., at 97-98. The trial judge need not review prosecutorial conduct in relation to other venires in other cases.
Having reviewed the requirements to prove a selective-prosecution claim, we turn to the showing necessary to obtain discovery in support of such a claim. If discovery is ordered, the Government must assemble from its own files documents which might corroborate or refute the defendant’s claim. Discovery thus imposes many of the costs present when the Government must respond to a prima facie case of selective prosecution. It will divert prosecutors’ resources and may disclose the Government’s prosecutorial strategy. The justifications for a rigorous standard for the elements of a selective-prosecution claim thus require a correspondingly rigorous standard for discovery in aid of such a claim.
The parties, and the Courts of Appeals which have considered the requisite showing to establish entitlement to discovery, describe this showing with a variety of phrases, like “colorable basis,” “substantial threshold showing,” Tr. of Oral Arg. 5, “substantial and concrete basis,” or “reasonable likelihood,” Brief for Respondents Martin et al. 30. However, the many labels for this showing conceal the degree of consensus about the evidence necessary to meet it. The Courts of Appeals “require some evidence tending to show the existence of the essential elements of the defense,” discriminatory effect and discriminatory intent. United States v. Berrios, 501 F. 2d 1207, 1211 (CA2 1974).
*469In this case we consider what evidence constitutes “some evidence tending to show the existence” of the discriminatory effect element. The Court of Appeals held that a defendant may establish a colorable basis for discriminatory effect without evidence that the Government has failed to prosecute others who are similarly situated to the defendant. 48 F. 3d, at 1516. We think it was mistaken in this view. The vast majority of the Courts of Appeals require the defendant to produce some evidence that similarly situated defendants of other races could have been prosecuted, but were not, and this requirement is consistent with our equal protection case law. United States v. Parham, 16 F. 3d 844, 846-847 (CA8 1994); United States v. Fares, 978 F. 2d 52, 59-60 (CA2 1992); United States v. Peete, 919 F. 2d 1168, 1176 (CA6 1990); C. E. Carlson, Inc. v. SEC, 859 F. 2d 1429, 1437-1438 (CA10 1988); United States v. Greenwood, 796 F. 2d 49, 52-53 (CA4 1986); United States v. Mitchell, 778 F. 2d 1271, 1277 (CA7 1985). As the three-judge panel explained, “ ‘[selective prosecution’ implies that a selection has taken place.” 21 F. 3d, at 1436.3
The Court of Appeals reached its decision in part because it started “with the presumption that people of all races commit all types of crimes — not with the premise that any type of crime is the exclusive province of any particular racial or ethnic group.” 48 F. 3d, at 1516-1517. It cited no authority for this proposition, which seems contradicted by the most recent statistics of the United States Sentencing Commission. Those statistics show: More than 90% of the persons sentenced in 1994 for crack cocaine trafficking were black, United States Sentencing Comm’n, 1994 Annual Report 107 (Table 45); 93.4% of convicted LSD dealers were white, ibid.; and 91% of those convicted for pornography or prostitution were white, id., at 41 (Table 13). Presumptions *470at war with presumably reliable statistics have no proper place in the analysis of this issue.
The Court of Appeals also expressed concern about the “evidentiary obstacles defendants face.” 48 F. 3d, at 1514. But all of its sister Circuits that have confronted the issue have required that defendants produce some evidence of differential treatment of similarly situated members of other races or protected classes. In the present case, if the claim of selective prosecution were well founded, it should not have been an insuperable task to prove that persons of other races were being treated differently than respondents. For instance, respondents could have investigated whether similarly situated persons of other races were prosecuted by the State of California and were known to federal law enforcement officers, but were not prosecuted in federal court. We think the required threshold — a credible showing of different treatment of similarly situated persons — adequately balances the Government’s interest in vigorous prosecution and the defendant’s interest in avoiding selective prosecution.
In the case before us, respondents’ “study” did not constitute “some evidence tending to show the existence of the essential elements of” a selective-prosecution claim. Berrios, supra, at 1211. The study failed to identify individuals who were not black and could have been prosecuted for the offenses for which respondents were charged, but were not so prosecuted. This omission was not remedied by respondents’ evidence in opposition to the Government’s motion for reconsideration. The newspaper article, which discussed the discriminatory effect of federal drug sentencing laws, was not relevant to an allegation of discrimination in decisions to prosecute. Respondents’ affidavits, which recounted one attorney’s conversation with a drug treatment center employee and the experience of another attorney defending drug prosecutions in state court, recounted hearsay and reported personal conclusions based on anecdotal evidence. The judgment of the Court of Appeals is therefore *471reversed, and the case is remanded for proceedings consistent with this opinion.
It is so ordered.
Other defendants had introduced this study in support of similar discovery motions in at least two other Central District cocaine prosecutions. App. 83. Both motions were denied. One District Judge explained from the bench that the 23-person sample before him was “statistically insignificant,” and that the evidence did not indicate “whether there is a bias in the distribution of crime that says black people use crack cocaine, his-panic people use powdered cocaine, Caucasian people use whatever it is they use.” Id., at 119, 120.
We have never determined whether dismissal of the indictment, or some other sanction, is the proper remedy if a court determines that a defendant has been the victim of prosecution on the basis of his race. Here, “it was the government itself that suggested dismissal of the indictments to the district court so that an appeal might lie.” 48 F. 3d 1508, 1510 (CA9 1995).
We reserve the question whether a defendant must satisfy the similarly situated requirement in a case “involving direct admissions by [prosecutors] of discriminatory purpose.” Brief for United States 15.