OPINION OF THE COURT
Chief Judge Kaye.The primary question before us is whether defendant, Educational Testing Service (ETS), a standardized testing firm, complied with procedures specified in its contract with high school senior Brian Dalton in refusing to release Dalton’s Scholastic Aptitude Test (SAT) score. Because the factual findings underlying the trial court’s determination that ETS failed to act in good faith in following those procedures were affirmed by the Appellate Division, have support in the record and are consequently beyond the scope of our review, we conclude — as did the trial court and Appellate Division — that ETS breached its contract with Dalton. Though we agree, moreover, with the *387courts below that specific performance is the appropriate remedy, we nevertheless conclude that the promised performance was good-faith compliance with the stated procedures, not release of the questioned scores as ordered by those courts.
I
In May 1991, Brian Dalton took the SAT, which was administered by ETS, at Holy Cross High School in Queens where Dalton was a junior. Six months later, in November, he took the examination a second time, as a senior, this time at John Bowne High School in Queens, and his combined score increased 410 points.
Because Dalton’s score increased by more than 350 points, his test results fell within the ETS category of "Large Score Differences” or "discrepant scores.” In accordance with ETS policy, members of the ETS Test Security Office therefore reviewed his May and November answer sheets. Upon a finding of disparate handwriting, the answer sheets were submitted to a document examiner, who opined that they were completed by separate individuals. Dalton’s case was then forwarded to the Board of Review, which preliminarily decided that substantial evidence supported cancelling Dalton’s November score.
Upon registering for the November SAT, Dalton had signed a statement agreeing to the conditions in the New York State edition of the Registration Bulletin, which reserved to ETS "the right to cancel any test score * * * if ETS believes that there is reason to question the score’s validity.” The Registration Bulletin further provided that, if "the validity of a test score is questioned because it may have been obtained unfairly, ETS [will] notifly] the test taker of the reasons for questioning the score” and offer the test-taker the following five options: (1) the opportunity to provide additional information, (2) confirmation of the score by taking a free retest, (3) authorization for ETS to cancel the score and refund all fees, (4) third-party review by any institution receiving the test score or (5) arbitration.
As specified in the Registration Bulletin, ETS apprised Dalton of its preliminary decision to cancel his November SAT score in a letter from Test Security Specialist Celeste M. Eppinger. Noting the handwriting disparity and the substantial difference between his May and November test results, Eppinger informed Dalton that "[t]he evidence suggests that *388someone else may have completed your answer sheet and that the questioned scores may be invalid.” She advised him that he could supply "any additional information that will help explain” this or, alternatively, elect one of the other options.
Eppinger enclosed the Procedures for Questioned Scores pamphlet with her letter, which reiterated the test-taker’s right to "submit additional relevant information” to the Board of Review supporting the validity of questioned scores. In cautioning test-takers to provide only information "relevant to the questions being raised,” the Procedures for Questioned Scores explained, "[f]or example, character references or testimonial letters do not explain handwriting differences.” As to the four additional options, the guide further explained, "ETS also offers other options * * * if additional information doesn’t resolve the questions about the validity of the scores. The option to provide additional information to resolve these questions may be used in combination with one or more of the[se] options.”
Dalton opted to present additional information to the Board of Review, including the following: verification that he was suffering from mononucleosis during the May examination; diagnostic test results from a preparatory course he took prior to the November examination (he had taken no similar course prior to the May SAT) that were consistent with his performance on that test; a statement from an ETS proctor who remembered Dalton’s presence during the November examination; and statements from two students — one previously unacquainted with Dalton — that he had been in the classroom during that test. Dalton further provided ETS with a report from a document examiner obtained by his family who concluded that Dalton was the author of both sets of answer sheets.
ETS, after several Board of Review meetings, submitted the various handwriting exemplars to a second document examiner who, like its first, opined that the May and November tests were not completed by the same individual. As a result, ETS continued to question the validity of Dalton’s November score.
At this point plaintiff Peter Dalton, father and natural guardian of Brian Dalton, filed a CPLR article 78 proceeding, later converted to an action at law, to prohibit ETS from cancelling Dalton’s November SAT score and to compel immediate release of the score. Following a 12-day nonjury trial, the trial court found that ETS failed "to make even rudimentary efforts to evaluate or investigate the information” furnished by *389Dalton and thus concluded that ETS failed to act in good faith in determining the legitimacy of Dalton’s score, thereby breaching its contract (155 Mise 2d 214, 225). The trial court premised this conclusion on its determination that the ETS Board of Review members failed to evaluate the information submitted because they believed Dalton’s presence at the November SAT to be wholly irrelevant to the handwriting issue and that he could controvert the Board’s preliminary finding that the score was invalid solely by taking a retest. As a remedy for the contractual breach, the trial court ordered ETS to release the November SAT score.
The Appellate Division affirmed. It too found that ETS ignored the documentation provided by Dalton and considered only the reports of its own document examiners. Like the trial court, the Appellate Division concluded that this failure to evaluate as well as to investigate Dalton’s information constituted a breach of contract. In light of these factual determinations, we agree that ETS breached its contract with Dalton but differ as to the scope of the relief.
II
By accepting ETS’ standardized form agreement when he registered for the November SAT, Dalton entered into a contract with ETS (see, AEB & Assocs. Design Group v Tonka Corp., 853 F Supp 724, 732). Implicit in all contracts is a covenant of good faith and fair dealing in the course of contract performance (see, Van Valkenburgh, Nooger & Neville v Hayden Publ. Co., 30 NY2d 34, 45, cert denied 409 US 875).
Encompassed within the implied obligation of each promisor to exercise good faith are " 'any promises which a reasonable person in the position of the promisee would be justified in understanding were included’ ” (Rowe v Great Atl. & Pac. Tea Co., 46 NY2d 62, 69, quoting 5 Williston, Contracts § 1293, at 3682 [rev ed 1937]). This embraces a pledge that "neither party shall do anything which will have the effect of destroying or injuring the right of the other party to receive the fruits of the contract” (Kirke La Shelle Co. v Armstrong Co., 263 NY 79, 87). Where the contract contemplates the exercise of discretion, this pledge includes a promise not to act arbitrarily or irrationally in exercising that discretion (see, Tedeschi v Wagner Coll., 49 NY2d 652, 659). The duty of good faith and fair dealing, however, is not without limits, and no obligation can be implied that "would be inconsistent with other terms of the contractual relationship” (Murphy v American Home Prods. Corp., 58 NY2d 293, 304).
*390The parties here agreed to the provisions in the Registration Bulletin, which expressly permit cancellation of a test score so long as ETS found "reason to question” its validity after offering the test-taker the five specified options. Nothing in the contract compelled ETS to prove that the test-taker cheated. Nor did the invitation to the test-taker to furnish ETS with relevant information reasonably and realistically translate into any requirement that ETS conduct a field investigation or gather evidence to verify or counter the test-taker’s documentation. Indeed, such an obligation would be inconsistent with the contractual language placing the burden squarely on the test-taker to overcome the ETS finding of score invalidity. ETS, therefore, was under no duty, express or implied, to initiate an external investigation into a questioned score.
The contract, however, did require that ETS consider any relevant material that Dalton supplied to the Board of Review. The Registration Bulletin explicitly afforded Dalton the option to provide ETS with relevant information upon notification that ETS questioned the legitimacy of his test score. Having elected to offer this option, it was certainly reasonable to expect that ETS would, at the very least, consider any relevant material submitted in reaching its final decision.
Dalton triggered this implied-in-law obligation on the part of ETS by exercising his contractual option to provide ETS with information (compare, Matter of Yaeger v Educational Testing Serv., 158 AD2d 602 [where test-taker declined to invoke any of the proffered options, ETS cancelled score in good faith and in accordance with terms of the contract]). Significantly, Dalton heeded the advice in the Procedures for Questioned Scores and tendered numerous documents that did more than simply deny allegations of wrongdoing or attest to his good character, such as medical evidence regarding his physical condition, statements by fellow test-takers, the statement of a classroom proctor and consistent diagnostic test results (compare, Swencki v Educational Testing Serv., No. C 81-0689 [WD KY] [test-taker sent letter to ETS explaining that he could not have cheated]; Matter of K. D. v Educational Testing Serv., 87 Misc 2d 657 [test-taker submitted sworn statement that he did not cheat]).
Nevertheless, with the exception of the document examiner’s report, ETS disputes the relevancy of this information. Specifically, ETS maintains that the sole issue before the Board of Review was the disparate handwriting and that evidence regarding Dalton’s health (apart from a damaged arm) or presence during both examinations is irrelevant to resolving that issue.
*391To be sure, the Procedures for Questioned Scores warned Dalton "to provide only additional information that is relevant to the questions being raised.” The Eppinger letter to Dalton, however, informed him that his November score was possibly invalid precisely because ETS believed "that someone else may have completed [his] answer sheet.” Thus, ETS expressly framed the dispositive question as one of suspected impersonation. Because the statements from the classroom proctor and November test-takers corroborated Dalton’s contention that he was present at and in fact took the November examination, they were relevant to this issue.
Likewise, inasmuch as the medical documentation concerning Dalton’s health at the time of the May SAT provided an explanation for his poor performance on that examination, and the consistent diagnostic test results demonstrated his ability to achieve such a dramatic score increase, these items were also germane to the question whether it was Dalton or an imposter who completed the November examination. Indeed, in its manual, Policies and Procedures Concerning Scores of Questionable Validity — which details internal ETS procedure regarding questioned scores — ETS offers several examples of "relevant information” that a test-taker might provide, including "a doctor’s report that the candidate was under the influence of medication at the time the low score was earned.” Regarding "a case of possible impersonation” in particular, the manual suggests that "other test results might demonstrate that the questioned score is not inconsistent with other measures of the candidate’s abilities.” Thus, Dalton’s material fell within ETS’ own definition of relevancy, as expressed in its manual and letter to Dalton.
The critical question then is whether the Board of Review made any effort to consider this relevant information submitted by Dalton. That is a factual inquiry. Both the trial court and the Appellate Division concluded that the Board utterly failed to evaluate the material. Given these affirmed findings, "our scope of review is narrow. This Court is without power to review findings of fact if such findings are supported by evidence in the record” (Humphrey v State of New York, 60 NY2d 742, 743).
Several Board of Review members — each member alone had the power to order release of Dalton’s November score — testified that they believed information establishing Dalton’s presence during the November examination to be irrelevant to *392their determination* and, moreover, that only a successful retest would validate Dalton’s score. Thus, there is support in the record for the factual determinations of the trial court and Appellate Division and they are binding on us. This is so notwithstanding inconsistent testimony by Board members that the Board did review Dalton’s information but found it unpersuasive. In light of the affirmed findings, the Court of Appeals simply does not have authority to weigh conflicting evidence and make its own factual determinations, as the dissent would do.
Consequently, this case is factually distinct from those relied upon by ETS, where the testing service considered but then rejected information provided by the test-taker (see, e.g., Langston v ACT, 890 F2d 380; Denburg v Educational Testing Serv., No. C-1715-83 [NJ Super Ct]; cf., Johnson v Educational Testing Serv., 754 F2d 20, 26, cert denied 472 US 1029 [noting that ETS provided test-taker with opportunity to be heard and to be represented by counsel]). When ETS fulfills its contractual obligation to consider relevant material provided by the test-taker and otherwise acts in good faith, the testing service — not the courts — must be the final arbiter of both the appropriate weight to accord that material and the validity of the test score. This Court will not interfere with that discretionary determination unless it is performed arbitrarily or irrationally.
Where, however, ETS refuses to exercise its discretion in the first instance by declining even to consider relevant material submitted by the test-taker, the legal question is whether this refusal breached an express or implied term of the contract, not whether it was arbitrary or irrational. Here, the courts below agreed that ETS did not consider the relevant informa*393tian furnished by Dalton. By doing so, ETS failed to comply in good faith with its own test security procedures, thereby breaching its contract with Dalton.
The dissent urges that because the trial court and Appellate Division relied in part on ETS’ failure to investigate Dalton’s information, they arguably employed an erroneous legal standard. Overlooked, however, is that both courts also concluded that ETS’ refusal to evaluate the material breached the contract with Dalton and, thus, employed a correct legal standard. Moreover, the crucial factual inquiry under the correct standard — whether ETS considered Dalton’s relevant material — has already been resolved by those courts. Because this factual finding dictates the legal conclusion that ETS breached the contract, remittal is unnecessary.
Ill
We agree with the trial court and Appellate Division that Dalton is entitled to specific performance of the contract. Dalton is not, however, entitled to release of his score as though fully validated. The goal of specific performance is to produce "as nearly as is practicable, the same effect as if the contract had been performed” (Farnsworth, Contracts § 12.5, at 823 [1982]). Had the contract here been performed, ETS would have considered the information provided by Dalton in reaching a final decision. ETS never promised to release a score believed to be invalid, and the validity of Dalton’s November SAT score has yet to be determined. Indeed, the trial court specifically noted that it was not resolving the question whether Dalton in fact took the November test.
In an analogous context, we have refused to compel a university to issue a degree to a student who had not fulfilled the academic requirements (see, Matter of Olsson v Board of Higher Educ., 49 NY2d 408). This reluctance to interfere with the exercise of academic discretion is motivated by sound considerations of public policy. "When an educational institution issues a diploma to one of its students, it is, in effect, certifying to society that the student possesses all of the knowledge and skills that are required by his [or her] chosen discipline” (id., at 413). Likewise, we have held that a college did not act arbitrarily in declining to "round off” a student’s failing grade so that she could graduate (see, Matter of McIntosh v Borough of Manhattan Community Coll., 78 AD2d 839, affd 55 NY2d 913).
The comparison between ETS and academic institutions is surely not exact, inasmuch as judicial restraint in matters of *394academic achievement is based, in part, on the inherently subjective nature of the evaluation to be made by professional educators (see, Tedeschi v Wagner Coll., 49 NY2d 652, 658, supra). Still, similar policy concerns militate against directing ETS to release a questioned score. When a standardized" testing service reports a score, it certifies to the world that the test-taker possesses the requisite knowledge and skills to achieve the particular score. Like academic credentials, if courts were to require testing services to release questioned scores, "the value of these credentials from the point of view of society would be seriously undermined” (Olsson, supra, at 413). Given the reliance that students, educational institutions, prospective employers and others place on the legitimacy of scores released by ETS, requiring challenged scores to be reported would be contrary to the public interest and exceed the scope of ETS’ promised performance.
While courts as a matter of policy are reluctant to intrude upon academic discretion in educational matters, they stand ready as a matter of law and equity to enforce contract rights. Where a contract is breached, moreover, and the injured party is entitled to specific performance, the remedy must be a real one, not an exercise in futility.
Dalton is entitled to relief that comports with ETS’ contractual promise — good-faith consideration of the material he submitted to ETS. We cannot agree with Dalton’s assumption that ETS will merely rubber-stamp its prior determination without good-faith attention to his documentation and that reconsideration by ETS will be an empty exercise. Our conclusion that the contract affords Dalton a meaningful remedy rests also on the provision in the Procedures for Questioned Scores allowing Dalton to utilize one or more of the remaining four options in combination with renewed consideration by the Board of Review. Those options — including third-party review by any institution receiving the test score as well as arbitration — remain available should ETS determine that the information submitted fails to resolve its concerns about the validity of the November score.
Accordingly, the Appellate Division order should be modified in accordance with this opinion and, as so modified, affirmed, without costs.
For example, when Shirley Kane-Orr — chairperson of the Board of Review and a member of the three-member panel of the Board that initially reviewed Dalton’s case on December 11,1991, the panel that again considered his case on January 3d, 1992, and the final panel to review Dalton’s case on February 7, 1992 — was asked whether she "or any member of the panel or the test security office ever questioned] [Dalton’s] presence in the classroom,” she answered, "[n]o”; when further asked by Dalton’s counsel, "[s]o am I to assume * * * that since you didn’t question his presence in the classroom, that it was not an issue in this case,” she responded, "[i]t was a non-issue in the sense that [it] was not an issue at all to be considered whether or not he was in the classroom.” Likewise, Sydell Carlton, member of the panel that considered Dalton’s case on January 17, 1992, was also asked whether she ever questioned Dalton’s presence in the classroom, to which she replied, "[t]hat was not an issue for us to decide. The issue for us to decide was, why are those handwritings disparate? His presence in the room was not, for us, in issue.”