Justice Blackmun
delivered the opinion of the Court.
r-H
In August 1983, petitioner Wayne T. Schmuck, a used-car distributor, was indicted in the United States District Court for the Western District of Wisconsin on 12 counts of mail fraud, in violation of 18 U. S. C. §§ 1341 and 1342. App. 3.
The alleged fraud was a common and straightforward one. Schmuck purchased used cars, rolled back their odometers, and then sold the automobiles to Wisconsin retail dealers for prices artificially inflated because of the low-mileage readings. These unwitting car dealers, relying on the altered odometer figures, then resold the cars to customers, who in turn paid prices reflecting Schmuck’s fraud. To complete the resale of each automobile, the dealer who purchased it from Schmuck would submit a title-application form to the Wisconsin Department of Transportation on behalf of his retail customer. The receipt of a Wisconsin title was a prerequisite for completing the resale; without it, the dealer could not transfer title to the customer and the customer could not obtain Wisconsin tags. The submission of the title-application form supplied the mailing element of each of the alleged mail frauds.
Before trial, Schmuck moved to dismiss the indictment on the ground that the mailings at issue — the submissions of the title-application forms by the automobile dealers — were not in furtherance of the fraudulent scheme and, thus, did not [708]*708satisfy the mailing element of the crime of mail fraud. Schmuck also moved under Federal Rule of Criminal Procedure 31(c)1 for a jury instruction on the then misdemeanor offense of tampering with an odometer, 15 U. S. C. §§ 1984 and 1990c(a) (1982 ed.).2 The District Court denied both motions.3 After trial, the jury returned guilty verdicts on all 12 counts.
A divided panel of the United States Court of Appeals for the Seventh Circuit reversed and remanded the case for a new trial. 776 F. 2d 1368 (1985). Although the panel rejected Schmuck’s claim that he was entitled to a judgment of acquittal because the mailings were not made in furtherance of his scheme, it ruled that under Rule 31(c) the District Court should have instructed the jury on the lesser offense of odometer tampering. The panel applied the so-called “inherent relationship” test for determining what constitutes a lesser included offense for the purpose of Rule 31(c). See, e. g., United States v. Whitaker, 144 U. S. App. D. C. 344, 349, 447 F. 2d 314, 319 (1971). Under that test, one offense is included in another when the facts as alleged in the indictment and proved at trial support the inference that [709]*709the defendant committed the less serious offense, and an “inherent relationship” exists between the two offenses. This relationship arises when the two offenses relate to the protection of the same interests and the proof of the greater offense can generally be expected to require proof of the lesser offense. Ibid. Applying this test, the court concluded that both the mail fraud and odometer tampering statutes protect against fraud, and that the proof of mail fraud generally entails proving the underlying fraudulent conduct.4 The panel then held that Schmuck was entitled to the lesser offense instruction because a rational jury could have found him guilty of odometer tampering, yet acquitted him of mail fraud on the ground that the mailings were too tangential to the fraudulent scheme to satisfy the requirements of mail fraud.
The Court of Appeals vacated the panel decision and ordered the case to be reheard en banc. 784 F. 2d 846 (1986). On rehearing, by a divided vote, 840 F. 2d 384 (1988), the en banc court rejected the “inherent relationship” test for defining lesser included offenses, and adopted instead the “elements test” whereby one offense is necessarily included within another only when the elements of the lesser offense form a subset of the elements of the offense charged. Id., at 387. The Court of Appeals found that the elements test “is grounded in the terms and history of Rule 31(c), comports with the constitutional requirement of notice to defendant of the potential for conviction of an offense not separately charged, permits a greater degree of certainty in the application of Rule 31(c), and harmonizes the concept of ‘necessarily included’ under Rule 31(c) with that of a lesser included offense where the issue is double jeopardy.” Id., at 388. Applying the elements test, the Court of Appeals held that Schmuck was not entitled to a jury instruction on the offense of odometer tampering because he could have been convicted [710]*710of mail fraud without a showing that he actually altered the odometers, but could not have been convicted of odometer tampering absent such a showing. Since the elements of odometer tampering are not a subset of the elements of mail fraud, odometer tampering did not qualify as a lesser included offense of mail fraud and, accordingly, the District Court was not required under Rule 31(c) to instruct the jury on the odometer-tampering offense.
We granted certiorari, 486 U. S. 1004 (1988), to define further the scope of the mail fraud statute and to resolve a conflict among the Circuits over which test to apply in determining what constitutes a lesser included offense for the purposes of Rule 31(c).5
II
“The federal mail fraud statute does not purport to reach all frauds, but only those limited instances in which the use of the mails is a part of the execution of the fraud, leaving all other cases to be dealt with by appropriate state law.” Kann v. United States, 323 U. S. 88, 95 (1944).6 To be part of the execution of the fraud, however, the use of the mails need not be an essential element of the scheme. Pereira v. United States, 347 U. S. 1, 8 (1954). It is sufficient for the [711]*711mailing to be “incident to an essential part of the scheme,” ibid., or “a step in [the] plot,” Badders v. United States, 240 U. S. 391, 394 (1916).
Schmuck, relying principally on this Court's decisions in Kann, supra, Parr v. United States, 363 U. S. 370 (1960), and United States v. Maze, 414 U. S. 395 (1974), argues that mail fraud can be predicated only on a mailing that affirmatively assists the perpetrator in carrying out his fraudulent scheme. The mailing element of the offense, he contends, cannot be satisfied by a mailing, such as those at issue here, that is routine and innocent in and of itself, and that, far from furthering the execution of the fraud, occurs after the fraud has come to fruition, is merely tangentially related to the fraud, and is counterproductive in that it creates a “paper trail” from which the fraud may be discovered. Brief for Petitioner 20-24. We disagree both with this characterization of the mailings in the present case and with this description of the applicable law.
We begin by considering the scope of Schmuck’s fraudulent scheme. Schmuck was charged with devising and executing a scheme to defraud Wisconsin retail automobile customers who based their decisions to purchase certain automobiles at least in part on the low-mileage readings provided by the tampered odometers. This was a fairly large-scale operation. Evidence at trial indicated that Schmuck had employed a man known only as “Fred” to turn back the odometers on about 150 different cars. Tr. 102-103. Schmuck then marketed these cars to a number of dealers, several of whom he dealt with on a consistent basis over a period of about 15 years. Id., at 33-34, 53. Indeed, of the 12 automobiles that are the subject of the counts of the indictment, 5 were sold to “P and A Sales,” and 4 to “Southside Auto.” App. 6-7. Thus, Schmuck’s was not a “one-shot” operation in which he sold a single car to an isolated dealer. His was an ongoing fraudulent venture. A rational jury could have concluded that the success of Schmuck’s venture de[712]*712pended upon his continued harmonious relations with, and good reputation among, retail dealers, which in turn required the smooth flow of cars from the dealers to their Wisconsin customers.
Under these circumstances, we believe that a rational jury could have found that the title-registration mailings were part of the execution of the fraudulent scheme, a scheme which did not reach fruition until the retail dealers resold the cars and effected transfers of title. Schmuck’s scheme would have come to an abrupt halt if the dealers either had lost faith in Schmuck or had not been able to resell the cars obtained from him. These resales and Schmuck’s relationships with the retail dealers naturally depended on the successful passage of title among the various parties. Thus, although the registration-form mailings may not have contributed directly to the duping of either the retail dealers or the customers, they were necessary to the passage of title, which in turn was essential to the perpetuation of Schmuck’s scheme. As noted earlier, a mailing that is “incident to an essential part of the scheme,” Pereira, 347 U. S'., at 8, satisfies the mailing element of the mail fraud offense. The mailings here fit this description. See, e. g., United States v. Locklear, 829 F. 2d 1314, 1318-1319 (CA4 1987) (retail customers obtaining title documents through the mail furthers execution of wholesaler’s odometer tampering scheme); United States v. Galloway, 664 F. 2d 161, 163-165 (CA7 1981) (same), cert. denied, 456 U. S. 1006 (1982); cf. United States v. Shryock, 537 F. 2d 207, 208-209 (CA5 1976) (local motor vehicle department’s mailing of title applications to state headquarters furthers retailer’s odometer-tampering scheme), cert. denied, 429 U. S. 1100 (1977).
Once the full flavor of Schmuck’s scheme is appreciated, the critical distinctions between this case and the three cases in which this Court has delimited the reach of the mail fraud statute — Kann, Parr, and Maze — are readily apparent. The defendants in Kann were corporate officers and directors [713]*713accused of setting up a dummy corporation through which to divert profits into their own pockets. As part of this fraudulent scheme, the defendants caused the corporation to issue two checks payable to them. The defendants cashed these checks at local banks, which then mailed the checks to the drawee banks for collection. This Court held that the mailing of the cashed checks to the drawee banks could not supply the mailing element of the mail fraud charges. The defendants’ fraudulent scheme had reached fruition. “It was immaterial to them, or to any consummation of the scheme, how the bank which paid or credited the check would collect from the drawee bank.” 323 U. S., at 94.
In Parr, several defendants were charged, inter alia, with having fraudulently obtained gasoline and a variety of other products and services through the unauthorized use of a credit card issued to the school district which employed them. The mailing element of the mail fraud charges in Parr was purportedly satisfied when the oil company which issued the credit card mailed invoices to the school district for payment, and when the district mailed payment in the form of a check. Relying on Kann, this Court held that these mailings were not in execution of the scheme as required by the statute because it was immaterial to the defendants how the oil company went about collecting its payment. 363 U. S., at 393.7
[714]*714Later, in Maze, the defendant allegedly stole his roommate’s credit card, headed south on a winter jaunt, and obtained food and lodging at motels along the route by placing the charges on the stolen card. The mailing element of the mail fraud charge was supplied by the fact that the defendant knew that each motel proprietor would mail an invoice to the bank that had issued the credit card, which in turn would mail a bill to the card owner for payment. The Court found that these mailings could not support mail fraud charges because the defendant’s scheme had reached fruition when he checked out of each motel. The success of his scheme in no way depended on the mailings; they merely determined which of his victims would ultimately bear the loss. 414 U. S., at 402.
The title-registration mailings at issue here served a function different from the mailings in Kann, Parr, and Maze. The intrabank mailings in Kann and the credit card invoice mailings in Parr and Maze involved little more than post-fraud accounting among the potential victims of the various schemes, and the long-term success of the fraud did not turn on which of the potential victims bore the ultimate loss. Here, in contrast, a jury rationally could have found that Schmuck by no means was indifferent to the fact of who bore the loss. The mailing of the title-registration forms was an essential step in the successful passage of title to the retail purchasers. Moreover, a failure of this passage of title would have jeopardized Schmuck’s relationship of trust and goodwill with the retail dealers upon whose unwitting cooperation his scheme depended. Schmuck’s reliance on our prior cases limiting the reach of the mail fraud statute is simply misplaced.
To the extent that Schmuck would draw from these previous cases a general rule that routine mailings that are in[715]*715nocent in themselves cannot supply the mailing element of the mail fraud offense, he misapprehends this Court’s precedents. In Parr the Court specifically acknowledged that “innocent” mailings — ones that contain no false information— may supply the mailing element. 363 U. S., at 390. In other cases, the Court has found the elements of mail fraud to be satisfied where the mailings have been routine. See, e. g., Carpenter v. United States, 484 U. S. 19, 28 (1987) (mailing newspapers).
We also reject Schmuck’s contention that mailings that someday may contribute to the uncovering of a fraudulent scheme cannot supply the mailing element of the mail fraud offense. The relevant question at all times is whether the mailing is part of the execution of the scheme as conceived by the perpetrator at the time, regardless of whether the mailing later, through hindsight, may prove to have been counterproductive and return to haunt the perpetrator of the fraud. The mail fraud statute includes no guarantee that the use of the mails for the purpose of executing a fraudulent scheme will be risk free. Those who use the mails to defraud proceed at their peril.
For these reasons, we agree with the Court of Appeals that the mailings in this case satisfy the mailing element of the mail fraud offenses.
Ill
Federal Rule of Criminal Procedure 31(c) provides in relevant part: “The defendant may be found guilty of an offense necessarily included in the offense charged.” As noted above, the Courts of Appeals have adopted different tests to determine when, under this Rule, a defendant is entitled to a lesser included offense instruction. The Seventh Circuit’s original panel opinion applied the “inherent relationship” approach formulated in United States v. Whitaker, 144 U. S. App. D. C. 344, 447 F. 2d 314 (1971):
“[DJefendant is entitled to invoke Rule 31(c) when a lesser offense is established by the evidence adduced at [716]*716trial in proof of the greater offense, with the caveat that there must also be an ‘inherent’ relationship between the greater and lesser offenses, i. e., they must relate to the protection of the same interests, and must be so related that in the general nature of these crimes, though not necessarily invariably, proof of the lesser offense is necessarily presented as part of the showing of the commission of the greater offense.” Id., at 349, 447 F. 2d, at 319.
The en banc Seventh Circuit rejected this approach in favor of the “traditional,” or “elements” test. Under this test, one offense is not “necessarily included” in another unless the elements of the lesser offense are a subset of the elements of the charged offense. Where the lesser offense requires an element not required for the greater offense, no instruction is to be given under Rule 31(c).
We now adopt the elements approach to Rule 31(c). As the Court of Appeals noted, this approach is grounded in the language and history of the Rule and provides for greater certainty in its application. It, moreover, is consistent with past decisions of this Court which, though not specifically endorsing a particular test, employed the elements approach in cases involving lesser included offense instructions.8
First, the wording of Rule 31(c), although not conclusive, supports the application of the elements approach. The Rule speaks in terms of an offense that is “necessarily included in the offense charged.” This language suggests that the comparison to be drawn is between offenses. Since offenses are statutorily defined, that comparison is appropriately conducted by reference to the statutory elements of the offenses in question, and not, as the inherent relationship approach [717]*717would mandate, by reference to conduct proved at trial regardless of the statutory definitions. Furthermore, the language of Rule 31(c) speaks of the necessary inclusion of the lesser offense in the greater. While the elements test is true to this requirement, the inherent relationship approach dispenses with the required relationship of necessary inclusion: the inherent relationship approach permits a lesser included offense instruction even if the proof of one offense does not invariably require proof of the other as long as the two offenses serve the same legislative goals.
In addition, the inherent relationship approach, in practice, would require that Rule 31(c) be applied in a manner inconsistent with its language. The Rule provides that a defendant “may be found guilty” of a lesser included offense, without distinguishing between a request for jury instructions made by the Government and one made by the defendant. In other words, the language of the Rule suggests that a lesser included offense instruction is available in equal measure to the defense and to the prosecution.9 Yet, under the inherent relationship approach, such mutuality is impossible.
It is ancient doctrine of both the common law and of our Constitution that a defendant cannot be held to answer a charge not contained in the indictment brought against him. See Ex parte Bain, 121 U. S. 1, 10 (1887); Stirone v. United [718]*718States, 361 U. S. 212, 215-217 (1960); United States v. Miller, 471 U. S. 130, 140, 142-143 (1985). This stricture is based at least in part on the right of the defendant to notice of the charge brought against him. United States v. Whitaker, 144 U. S. App. D. C., at 350-351, 447 F. 2d, at 320-321. Were the prosecutor able to request an instruction on an offense whose elements were not charged in the indictment, this right to notice would be placed in jeopardy. Specifically, if, as mandated under the inherent relationship approach, the determination whether the offenses are sufficiently related to permit an instruction is delayed until all the evidence is developed at trial, the defendant may not have constitutionally sufficient notice to support a lesser included offense instruction requested by the prosecutor if the elements of that lesser offense are not part of the indictment. Accordingly, under the inherent relationship approach, the defendant, by in effect waiving his right to notice, may obtain a lesser offense instruction in circumstances where the constitutional restraint of notice to the defendant would prevent the prosecutor from seeking an identical instruction.10 The elements test, in contrast, permits lesser offense instructions only in those cases where the indictment contains the elements of both offenses and thereby gives notice to the defendant that he may be convicted on either charge. This approach preserves the mutuality implicit in the language of Rule 31(c).
Second, the history of Rule 31(c) supports the adoption of the elements approach. The Rule, which has not been amended since its adoption in 1944, is the most recent derivative of the common-law practice that permitted a jury to find a defendant “guilty of any lesser offense necessarily included in the offense charged.” Beck v. Alabama, 447 U. S. 625, [719]*719633 (1980). Over a century ago, Congress codified the common law for federal criminal trials, providing in the Act of June 1, 1872, ch. 255, '§ 9, 17 Stat. 198, that “in all criminal causes the defendant may be found guilty of any offence the commission of which is necessarily included in that with which he is charged in the indictment.” Rule 31(c) was intended to be a restatement of this “pre-existing law.” See Keeble v. United States, 412 U. S. 205, 208, n. 6 (1973). Accordingly, prevailing practice at the time of the Rule’s promulgation informs our understanding of its terms, and, specifically, its limitation of lesser included offenses to those “necessarily included in the offense charged.”
The nature of that prevailing practice is clear. In Giles v. United States, 144 F. 2d 860 (1944), decided just three months before the adoption of Rule 31(c), the Court of Appeals for the Ninth Circuit unequivocably applied the elements test to determine the propriety of a lesser included offense instruction: “ ‘To be necessarily included in the greater offense the lesser must be such that it is impossible to commit the greater without first having committed the lesser.’” Id., at 861, quoting House v. State, 186 Ind. 593, 595-596, 117 N. E. 647, 648 (1917). This approach, moreover, was applied consistently by state courts. Indeed, in State v. Henry, 98 Me. 561, 564, 57 A. 891, 892 (1904), the Supreme Judicial Court of Maine concluded that “a practically universal rule prevails, that the verdict may be for a lesser crime which is included in a greater charged in the indictment, the test being that the evidence required to establish the greater would prove the lesser offense as a necessary element.” The California Supreme Court in People v. Kerrick, 144 Cal. 46, 47, 77 P. 711, 712 (1904), stated: “To be ‘necessarily included’ in the offense charged, the lesser offense must not only be part of the greater in fact, but it must be embraced within the legal definition of the greater as a part thereof.” See also State v. Marshall, 206 Iowa 373, 375, 220 N. W. 106 (1928); People ex rel. Wachowicz v. Martin, 293 N. Y. 361, [720]*720364, 57 N. E. 2d 53, 54-55 (1944). This Court’s decision in Stevenson v. United States, 162 U. S. 313 (1896), reflects the “practically universal” practice. There, in holding that the defendant in a murder charge was entitled to a lesser included offense instruction on manslaughter under the statutory predecessor to Rule 31(c), the Court engaged in a careful comparison of the statutory elements of murder and manslaughter to determine if the latter was a lesser included offense of the former. Id., at 320. In short, the elements approach was settled doctrine at the time of the Rule’s promulgation and for more than two decades thereafter. In its restatement of “pre-existing law,” Keeble v. United States, 412 U. S., at 208, n. 6, Rule 31(c) incorporated this established practice.11
Third, the elements test is far more certain and predictable in its application than the inherent relationship approach. Because the elements approach involves a textual comparison of criminal statutes and does not depend on inferences that may be drawn from evidence introduced at trial, the elements approach permits both sides to know in advance what jury instructions will be available and to plan their trial strategies accordingly. The objective elements approach, moreover, promotes judicial economy by providing a clearer rule [721]*721of decision and by permitting appellate courts to decide whether jury instructions were wrongly refused without reviewing the entire evidentiary record for nuances of inference.
The inherent relationship approach, in contrast, is rife with the potential for confusion. Finding an inherent relationship between offenses requires a determination that the offenses protect the same interests and that “in general” proof of the lesser “necessarily” involves proof of the greater. In the present case, the Court of Appeals appropriately noted: “These new layers of analysis add to the uncertainty of the propriety of an instruction in a particular case: not only are there more issues to be resolved, but correct resolution involves questions of degree and judgment, with the attendant probability that the trial and appellate courts may differ.” 840 F. 2d, at 389-390. This uncertainty was illustrated here. The three judges of the original appellate panel split in their application of the inherent relationship test to the offenses of mail fraud and odometer tampering. 776 F. 2d, at 1373-1375 (opinion concurring in part and dissenting in part). In the context of rules of criminal procedure, where certainty and predictability are desired, we prefer the clearer standard for applying Rule 31(c).
IV
Turning to the facts of this case, we agree with the Court of Appeals that the elements of the offense of odometer tampering are not a subset of the elements of the crime of mail fraud. 840 F. 2d, at 386. There are two elements in mail fraud: (1) having devised or intending to devise a scheme to defraud (or to perform specified fraudulent acts), and (2) use of the mail for the purpose of executing, or attempting to execute, the scheme (or specified fraudulent acts). The offense of odometer tampering includes the element of knowingly and willfully causing an odometer to be altered. This element is not a subset of any element of mail fraud. Knowingly and willfully tampering with an odometer is not identical to devis[722]*722ing or intending to devise a fraudulent scheme. Compare 18 U. S. C. § 1341 with 15 U. S. C. §§ 1984 and 1990c(a).
V
We conclude that Schmuck’s conviction was consistent with the statutory definition of mail fraud and that he was not entitled to a lesser included offense instruction on odometer tampering. The judgment of the Court of Appeals, accordingly, is affirmed.
It is so ordered.