OPINION OF THE COURT
PER CURIAM:This appeal presents the question whether racial discrimination has been proved by a preponderance of the evidence.
Plaintiffs McNeil and Jones, who are black, filed their complaint in the United States District Court for the District of New Jersey charging their employer, the Division of Alcoholic Beverage Control (a unit of New Jersey state government), and the present and three former directors of the Division, with engaging in a pattern of racially discriminatory hiring and promotional procedures and harassment. The complaint sought declaratory, injunctive, and monetary relief pursuant to Title VII of the Civil Rights Act of 1964, 42 U.S.C. § 2000e-5(g) (1976), and 42 U.S.C. §§ 1981, 1983 & 1988 (1976).1 The case was tried to the court without a jury. Although plaintiffs established a “borderline” prima facie case of racial discrimination in promotions, the district court found that the defendants had rebutted the plaintiffs’ case sufficiently to warrant a judgment in defendants’ favor. The plaintiffs appeal and we affirm.
I.
As the district court observed, the Division is “essentially a police agency” whose principal unit is the Enforcement Bureau. Of the Division’s approximately 150 employees, more than half are’ assigned to the Enforcement Bureau. This Bureau, in turn, is divided into five sections: wholesale, retail, undercover, front, and adminis*180trative. For most of their tenure in the Division, the plaintiffs have served in the undercover section.2 Undercover agents visit “licensees,” i. e., businesses licensed to sell liquor under title 33 of the New Jersey Revised Statutes, to monitor compliance with the applicable state laws. Such agents are thus exposed not only to the personal risk typical in police work, but also, due partly to the “cash operation” characteristic of the retail liquor business, substantial temptations of bribery.3 The Bureau’s employees are “unclassified” and are not subject to civil service regulations. The Director, appointed by the Governor of the State, makes the appointments to the Bureau. Appointees are removable at will for the first three years of their employment and thereafter only for cause.
McNeil and Jones, both tenured employees, have each worked for the Enforcement Bureau since the mid-1950s. They have yet to rise to the level of supervisor, although each has served as a fill-in supervisor and Jones has received supervisor training. Although McNeil had been promoted once in 1964, and Jones had received a promotion in 1971 pursuant to a Division-wide reorganization, the only other promotions received by either plaintiff followed shortly after each had filed discrimination complaints against the Division. At trial, plaintiffs attempted to show that their failures to obtain promotion to the rank of supervisor had been the result of intentionally disparate treatment; their relative salary and grade within the ranks of the enforcement bureau agents had not improved to the degree experienced by some nonminority agents in the Bureau who possessed equal or less seniority.
Plaintiffs presented evidence of several incidents during their tenure that could be said to indicate that their nonminority superiors had harassed or refused to promote them for reasons of racial prejudice, finding an easy cover for illicit motives in the rather informal promotion system in place in the relevant time period. In connection with each such incident, however, there was also evidence that plaintiffs’ superiors were motivated by factors other than racial prejudice. This evidence was buttressed by evidence that one other black agent had been promoted to the supervisory grade; that some nonminority agents with seniority equal to or greater than plaintiffs had also remained at or below plaintiffs’ grade; and that plaintiffs never had taken the steps necessary to be considered for promotion to the rank of supervisor.4
II.
The latitude available to us on this appeal is extremely narrow. The plaintiffs ask us to set aside the factual finding that plaintiffs’ failure to advance was not due to intentional racial discrimination. They do not take issue with the legal rulings of the trial court. But the trial court’s findings of fact may not be set aside “unless clearly erroneous, and due regard shall be given to the opportunity of the trial court to judge of the credibility of the witnesses.” Fed.R. Civ.P. 52(a). We have reviewed the record and find no basis for concluding that the findings of the trial court were clearly erroneous.
It would be impermissible for us to decide in the first instance whether to credit the testimony upon which the trial judge predicated his findings. In this “disparate treatment” case, the trial court must be upheld so long as this court holds that if the testimony were true, the finder of fact would be entitled to infer the presence of legitimate job-related reasons for defendants’ acts and an absence of discriminatory motive. See generally Texas Department of Community Affairs v. Burdine, - U.S. -, 101 S.Ct. 1089, 67 L.Ed.2d 207 (1981); *181McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973). Plaintiffs do not argue that the motives put forward by the defendants, if true, would be legally insufficient to rebut the prima facie case made out by the plaintiffs. Eather, they invite us to find that inferences of bad motive must be drawn from the evidence. However, the trial court found no evidence tending to show that the defendants’ articulation of legitimate reasons constituted a “pretext to conceal an underlying discrimination.”
Perhaps one of the most damning incidents put forward by the plaintiffs tended to show that Inspector Dalton, while ostensibly engaged in a good-faith investigation of alleged bribery within the Division, was in fact “out to get” McNeil and Jones, simply because the latter were black. However, nothing came of Dalton’s alleged illicit plan; to the contrary, the plaintiffs were given a clean bill of health and subsequently then-Director Bower offered them what the district court interpreted as an opportunity to advance themselves within the Division by taking on a special assignment.5 Despite plaintiffs’ vigorous claims to the contrary, we cannot conclude that Judge Biunno’s inference with regard to the offered assignment was clearly wrong. In a typical employment relationship, it is just such a willingness to shoulder an additional assignment that brings the assignee future rewards, should he perform well. The plaintiffs raise a plausible argument that their advancement was not the motive behind this particular assignment. But their argument does not rise to the level of making the alternate conclusion unreasonable. Giving due regard to the trial judge’s opportunity to view the witnesses and form an opinion as to their veracity, we cannot say that his conclusion was clear error.6
Plaintiffs also argue that McNeil’s forced transfer from undercover to retail in 1971 constituted a discriminatory refusal to promote. At approximately the same time as McNeil’s transfer, a Division-wide reorganization resulted in the promotion of all undercover agents, including Jones. As a result of his transfer to retail, however, McNeil remained at his pre-reorganization salary. Former Director McDonough testified that McNeil’s transfer was the result of McNeil’s receipt of publicity in connection with a city election campaign. As McDonough saw it, open publicity was inconsistent with effective performance of undercover work. In the retail section, on the other hand, “cover” was not a factor. The district court concluded that the reason for McNeil’s transfer was “rationally related to the agency function, i. e., a valid ‘business reason,’ equally true without regard to race, sex, and the like.”
*182As in the case of the Dalton incident, we cannot conclude that the district court’s decision to credit McDonough’s professed nondiscriminatory motive was clear error.7 Nor does the record provide sufficient evidence to convince us that the district court committed clear error when he credited those witnesses who testified that the other incidents of which plaintiffs complain arose out of legitimate job-related, rather than race-related, concerns.
Had this been a jury trial, the plaintiffs would have retained the ultimate burden of persuasion even assuming that the defendants’ proffered justifications had not persuaded the court that they constituted the defendants’ true motivations. All that was necessary was a clear statement of a legitimate, nondiseriminatory reason, set forth through the introduction of admissible evidence. “It is sufficient if the defendant’s evidence raises a genuine issue of fact as to whether it discriminated against the plaintiff.” Texas Department of Community Affairs v. Burdine, at-, 101 S.Ct. at 1094 (footnote omitted); see also Jackson v. U. S. Steel Corp., 624 F.2d 436 (3d Cir. 1980). Such a showing was made in this case. To reverse now would thus be to either substitute our own credibility determinations for those made by the trier of fact, or to impose on the defendants the burden of persuading the court that they “had convincing, objective reasons for preferring the chosen applicant^ for promotion] above the plaintiff[s].” At-, 101 S.Ct. at 1095 (footnote omitted). As Bur-dine indicates, the latter imposition is as impermissible as the former substitution.
We note that on this appeal plaintiffs have attempted to fuel our suspicions of disparate treatment by providing selected statistical data tending to show that blacks were relegated to an inferior status in the Division, perhaps through the facially neutral medium of the Division’s former highly subjective promotional procedures.8 However, at trial no demographic or other statistical evidence was introduced. Under these circumstances, this court can hardly undertake now to speculate on the statistical likelihood that plaintiffs’ failure to advance within the Division could or could not have been caused by factors other than racial discrimination.9
*183III.
In view of the evidence produced at trial and the limited scope of our review on this appeal, the judgment of the district court will be affirmed.
. Plaintiffs invoked jurisdiction of the district court pursuant to 42 U.S.C. § 2000e-5, 28 U.S.C. §§ 1331(a), 1343(3) & (4) (1976). Plaintiffs had previously received a right-to-sue letter from the Equal Employment Opportunity Commission on March 9, 1976.
. Since 1971, McNeil has served in the retail section.
. The trial court found that “the agency is aware of this risk and undertakes discharge proceedings in every case, without regard to race, sex, etc., where the evidence sustains the charge.”
. After having served as a supervisory fill-in, Jones actually let it be known that he preferred not to take on supervisory details.
. The special assignment called for the creation of an experimental “flying squad” of black agents who would work undercover in establishments catering to black patrons. McNeil was to be the squad supervisor, but his salary and job classification would not be immediately raised. Plaintiffs turned down the special assignment, perhaps because they distrusted Bower’s motives. On this appeal they contend that the proposal would have effectively removed the black agents from competition with whites for promotions. There is no suggestion, however, that the all-black composition of the squad was itself improper. On the contrary, plaintiffs argued here and at trial that the pairing of white and black agents in this type of undercover work was, in their own view, undesirable. “This practice endangered black agents when they went into establishments that were predominantly frequented by black patrons.” Brief for plaintiffs-appellants at 8.
The trial court found that Bower’s offer of the flying squad assignment was sincere and well-motivated. Director McDonough had previously organized a “flying squad” with agent Salvatore (a white man) at its head. Far from removing Salvatore from contention for promotion over those agents outside his squad, his successful performance as flying squad leader, found the district court, played a significant part in his later promotion. Plaintiffs’ unfortunate misunderstanding of Bower’s motive was caused by, if anything, a “breach of trust” on the part of Vestal, the agent assigned to investigate plaintiffs. Bower had no knowledge of this breach.
. As for the harm that Dalton apparently attempted to cause, he failed, and he was subsequently fired for accepting bribes. Thus no harm resulted to the plaintiffs and no lasting bias can be inferred from the incident.
. We find further support for the judge’s conclusion in the benefit Jones derived from the reorganization. There is no contention here that McNeil’s failure to benefit from the reorganization impermissibly infringed his first amendment rights, or was in any way connected to the issues — either race-related or otherwise — of McNeil’s candidacy.
. The defendants have offered their own statistical data, equally or more convincing than plaintiffs’, in rebuttal.
. See Miller v. Weber, 577 F.2d 75, 77 (8th Cir. 1978) (“It is the plaintiffs responsibility to produce a meaningful statistical comparison.”); cf. Hazelwood School District v. United States, 433 U.S. 299, 312, 97 S.Ct. 2736, 2744, 53 L.Ed.2d 768 (1977) (remanding because the Court could make appropriate use of statistical evidence only after consideration of “all the surrounding facts and circumstances” based on trial court findings); EEOC v. Greyhound Lines, Inc., 635 F.2d 188 (3d Cir. 1980) (rejecting a disparate impact claim because there was insufficient evidence that an employer’s policy excluded more blacks than whites from a job). True,
[sjometimes a clear pattern, unexplainable on grounds other than race, emerges from the effect of the state action even when the governing legislation appears neutral on its face.... But such cases are rare. Absent a pattern as stark as that in Gomillion [v. Lightfoot, 364 U.S. 339, 81 S.Ct. 125, 5 L.Ed.2d 110 (1960)] or Yick Wo [v. Hopkins, 118 U.S. 356, 6 S.Ct. 1064, 30 L.Ed. 220 (1886)], impact alone is not determinative, and the Court must look to other evidence.
Arlington Heights v. Metropolitan Housing Corp., 429 U.S. 252, 266, 97 S.Ct. 555, 563, 50 L.Ed.2d 450 (1977) (footnotes omitted); accord, Hazelwood School District v. United States, 433 U.S. at 307-08, 97 S.Ct. at 2741. Of course, in this case we deal neither with a disparate “impact” claim as in Greyhound, or a claim of discriminatory legislative purpose, as in Arlington Heights. However, it is clear that to the extent that an offer of statistical proof falls far short of demonstrating proscribed discrimination in cases of either of these types, it also must fail to demonstrate the sort of disparate treatment alleged here. See generally International Brotherhood of Teamsters v. United States, 431 U.S. 324, 335 n.15, 97 S.Ct. 1843, 1854 n.15, 52 L.Ed.2d 396 (1977). A fortiori, where as here plaintiffs make no offer of statistical proof in the district court, this court cannot use statistics to infer error on the part of the lower court.