OPINION OF THE COURT
SLOVITER, Circuit Judge.On March 20, 1971 the appellant, Curtis Worthy, was operating a 30 ton crane in the 18 Inch Hot Strip Mill of the United States Steel plant at Fairless Hills, Pennsylvania, when it collided with another crane operated by Raymond Weaver. As a result of the collision, Worthy was suspended for ten days and demoted from his position as crane operator, classified as Coiler, Class 12, to Bander, Class 4. Worthy, who is black, asserts that the suspension and the demotion constituted disparate treatment against him in violation of Title VII of the Civil Rights Act of 1964, as amended, 42 U.S.C. § 2000e et seq., and 42 U.S.C. § 1981. The district court found for the defendant employer (U.S. Steel) and Worthy appeals. The principal issue is whether the district court gave sufficient consideration to the employer’s treatment of white employees in determining that there was no discrimination.
On this appeal, the parties stress different facts and evidence in support of their opposing contentions. Our view of the case makes it unnecessary for us to undertake a detailed examination of the evidence. In essence, Worthy claims that he was given an unduly harsh ten-day suspension and demotion by U.S. Steel because he was black. U.S. Steel claims that these were valid safety and disciplinary measures of a nondiscriminatory nature.
Worthy’s case was consolidated for discovery and trial with a class action alleging racial discrimination at the plant. Trial of the consolidated action consumed 85 days. The class action suit involved a broad based attack on many aspects of the hiring and employment practices at the facility. The district court issued two lengthy opinions dealing primarily with the issues applicable to the class, one on July 25, 1977 at the conclusion of plaintiffs’ case, reported as Dickerson v. United States Steel Corp., 439 F.Supp. 55 (E.D.Pa.1977), and the second on August 2, 1978 at the conclusion of all the testimony, reported at 472 F.Supp. 1304 (E.D.Pa.1978). Of the three individual claims which were consolidated with the class action, Worthy’s was the only individual case, the other two being the claims of the class representatives. Although Worthy’s case took almost ten days to try, it was considered by the court in a brief discussion at the conclusion of both opinions. See 439 F.Supp. at 89-90; 472 F.Supp. at 1356.
In its first opinion the district court held that Worthy had sustained the burden of proving a prima facie case of disparate treatment discrimination because “[t]he testimony and documents submitted by the plaintiff show that whites were not so [demoted and suspended] under similar circumstances.” 439 F.Supp. at 90. In the final opinion at the conclusion of all of the testimony, the district court disposed of Worthy’s individual claim in a two paragraph portion of the larger opinion. The court held that Worthy’s suspension and demotion were nondiscriminatory when viewed against the handling of the white crane-men’s cases. The court determined that Worthy’s safety record of three accidents and other citations for improper crane operation “placed him at the same level, if not worse, as white cranemen who were demoted.” Since Worthy failed on many occasions to properly operate the crane and thus endangered the safety of other men, the court concluded that removing him from the craneman’s job, as had been done with whites, “was a rational and nondiscriminatory safety measure, not a discriminatory discipline.” In addition, the court found that the ten-day suspension was not excessive, since it was not only for the improper crane operation but for violating two additional rules — failing to report the accident and failing to cease operation of the crane after the accident. Therefore the ten-day “discipline was not unreasonable, even when compared with lesser suspensions re*701ceived by whites,” since the whites apparently did not violate the two other rules. 472 F.Supp. at 1356. On Worthy’s motion for reconsideration the court reaffirmed its finding that the demotion was not discriminatory.
On appeal, Worthy asserts that the district court’s conclusion that his demotion was a nondiscriminatory safety measure and that his ten-day suspension was not unreasonable or discriminatory was clearly erroneous because the record shows that whites with worse safety records were not demoted and that whites with comparable infractions were not disciplined. Worthy also claims that the district court erred in comparing his treatment with those white cranemen who were demoted for safety reasons rather than with those white cranemen with safety violations who were not demoted.
This court recently had occasion to review the principal Supreme Court decisions which set forth the appropriate procedure to be followed in determining whether plaintiff has proved discrimination in violation of Title VII.1 In Kunda v. Muhlenberg College, No. 79-1135, slip op. at 18 (3d Cir. Feb. 19, 1980), we noted that in cases alleging disparate treatment, when plaintiff produces sufficient evidence to establish a prima facie case, the burden shifts to defendant to articulate some legitimate nondiscriminatory reason for the employment action. The burden of proof remains with plaintiff who must, to prevail, prove that the reason articulated was pretextual.
Although cases are analyzed in terms of these three phases, there is no requirement that the evidence be introduced in such a compartmentalized form. Whack v. Peabody & Wind Engineering Co., 595 F.2d 190, 193 (3d Cir. 1979). Thus, it is possible that the plaintiff’s evidence relevant to the question of pretext can be presented as part of plaintiff’s initial evidence going to the prima facie case itself. It is also possible that one party’s evidence and proof may be developed in the context of the other party’s case, such as by cross-examination. Although we have never been technical about the order in which the evidence is developed, it is necessary that the trial court make clear findings of fact as to each of the essential phases of the case, so that they can be examined in light of the evidence on record and the applicable legal principles. H. Prang Trucking Co. v. Local Union No. 469, 613 F.2d 1235, 1239 (3d Cir. 1980); O’Neill v. United States, 411 F.2d 139 (3d Cir. 1969).
In this appeal the court’s finding that plaintiff had established a prima facie case is no longer in issue. Similarly, since the court found that defendant’s employment decisions were rational and nondiscriminatory safety measures, this subsumes a finding that defendant met the lesser burden of articulating a legitimate nondiscriminatory reason for the actions. There is no question that employment decisions may be made on the basis of safety considerations without violating the proscriptions of Title VIL “[T]he employer may rebut a prima facie case of the employee by showing that the maintenance of safety and efficiency requires the practice which obtains.” Muller v. United States Steel Corp., 509 F.2d 923, 928 (10th Cir.), cert. denied, 423 U.S. 825, 96 S.Ct. 39, 46 L.Ed.2d 41 (1975). Concern about the safety of coworkers or the general public may well be one of the leading legitimate non-discriminatory reasons that can be offered. See New York City Transit Authority v. Beazer, 440 U.S. 568, 587 n.31, 99 S.Ct. 1355, 1366 n.31, 59 L.Ed.2d 587 (1979). Patently, an employer responsible for life and property used in production must be given leeway of judgment as to the appropriate safeguard measures to employ.
While safety is a legitimate basis for an employment decision, a principle not *702challenged by the appellant in this case, the fact that safety considerations may have been involved does not in itself insulate the decision from a finding of discrimination. Appellant argues that at issue in this case are not the safety or disciplinary rules per se but their application in more severe fashion to black workers than to white workers. Plaintiff is entitled to demonstrate that a facially neutral decision was in fact pretext for discriminatory treatment. McDonnell Douglas Corp. v. Green, 411 U.S. 792, 804, 93 S.Ct. 1817, 1825, 36 L.Ed.2d 668 (1973). Particularly relevant to such a showing would be evidence that white employees with comparable records were not similarly treated. Id.
Appellant attempted to meet that burden in this case with evidence that white crane-men were treated less harshly. The trial court rejected appellant’s contention by finding that white cranemen of comparable safety records were also demoted. Although the court did not label its finding in terms of pretext, we view its conclusion as a finding that appellant failed to sustain his burden of showing that the safety reasons were pretextual in view of the demotion of similarly situated whites. Appellant has two arguments directed to this finding: first, that it was clearly erroneous to have found that the three whites who were demoted were comparable to plaintiff; and second, that the court’s finding of no pretext is legally insufficient because the court failed to consider evidence showing that there were comparable whites who were not demoted or disciplined.
Appellee argues that it is evident from the record that the demotion and disciplinary suspension imposed on Worthy must be viewed in the context of his entire safety record; that discipline was imposed on him in sequential fashion beginning with a warning and culminating in the ten-day suspension and demotion; and most significantly, that no other craneman had a safety record as serious as that of Worthy because he alone had three collisions in addition to other safety violations including dropping a block to the floor below. The claim that different treatment of one or more employees was justified because of their employment record was considered by the Supreme Court in McDonald v. Santa Fe Trail Transportation Co., 427 U.S. 273, 96 S.Ct. 2574, 49 L.Ed.2d 493 (1976). There the employer attempted to justify the dismissal of the plaintiffs, who were white, because of their commission of a serious criminal offense against their employer. The Court noted that this evidence would not be an acceptable reason if other culpable employees, in that case the allegedly favored blacks, were treated differently. The Court noted:
The Act prohibits all racial discrimination in employment, without exception for any group of particular employees, and while crime or other misconduct may be a legitimate basis for discharge, it is hardly one for racial discrimination. Indeed, the Title VII plaintiff in McDonnell Douglas had been convicted for a nontrivial offense against his former employer. It may be that theft of property entrusted to an employer for carriage is a more compelling basis for discharge than obstruction of an employer’s traffic arteries, but this does not diminish the illogic in retaining guilty employees of one color while discharging those of another color. (Footnotes omitted).
Id. at 283-84, 96 S.Ct. at 2580 (emphasis in original).
A relevant issue developed under the facts in this case is therefore whether “employees involved in acts against [the employer] of comparable seriousness were nevertheless retained . . ..” McDonnell Douglas Corp. v. Green, 411 U.S. at 804, 93 S.Ct. at 1825 (emphasis added). In Corley v. Jackson Police Department, 566 F.2d 994 (5th Cir. 1978), two black police officers who were discharged because they allegedly had accepted a bribe brought suit under Title VII claiming discrimination. The judgment of the trial court in favor of the employer was reversed on appeal because the court failed to give adequate consideration to treatment of other employees. The appellate court stated:
*703[T]he district court confused the issue of whether the employees actually committed the crime on which their discharge was based with the issue of whether the crime committed was used by the employer as a pretext for discrimination, that is, whether the employer applied the same criterion to other employees accused of the same or comparable offenses. The district court’s opinion addresses no evidence relevant to pretext, the most important of which indicated that, unlike Corley and Carter, other Jackson police officers named by Johnson and Strahan were neither investigated nor discharged. Under these circumstances, the pro forma conclusion of the district court that the discharges of Corley and Carter were “proper” and “for good cause” is insufficient to assure this Court on review that the proper legal standard regarding pretext was applied in reaching the ultimate conclusion. (Footnotes omitted).
Id. at 999-1000. Similarly, in Berbiglia, Inc. v. NLRB, 602 F.2d 839, 843-4 (8th Cir. 1979), the court held that although evidence that employees were padding their commissions would ordinarily justify their discharge, the fact that the employer accorded different treatment to similar activity by others demonstrated the discriminatory nature of the discharges which were directed toward union activists.
Appellee may well be correct that the facts which it cited placed Worthy in a separate category which justified his harsher treatment. In order to determine comparability, however, the district court was obliged to conduct a thorough inquiry into the treatment of comparable employees. That inquiry could not stop with the determination that comparable whites were demoted. Before a finding could be made that there was no pretext, it was necessary for the district court to consider also the evidence presented by Worthy purportedly showing that there were comparable whites who were not demoted or not disciplined. Otherwise, an employer involved in racial discrimination could demote or discipline some whites as a token gesture to show similar treatment while leaving the majority of the favored employees free from any disciplinary action.
We do not suggest that this was done in this case. We are unfortunately hampered by the lack of findings by the trial court on the relevant issues. Appellee concedes that the district court made no finding on the comparability of Worthy’s record to those white employees who were neither demoted nor disciplined for ten days. Counsel for the appellee suggested at oral argument that the district court’s failure to make specific findings on all of the relevant issues was attributable to the court’s preoccupation with the massive class action to which this individual case was tagged. We recognize the magnitude of the task faced by the district court, but nevertheless we cannot sit as a fact finder and cull through a record of over 15,000 transcript pages in order to determine whether those employees who were not demoted or disciplined should or should not be considered comparable to Worthy.
The difficulty faced by an appellate court when there have been inadequate findings of fact in a case claiming discriminatory treatment was noted by the Fifth Circuit:
[I]f we assume that the district court applied the McDonnell Douglas pretext test, we could only speculate as to whether it correctly applied that test. The district court made no findings of fact relevant to the pretext issue and it is not our function to make credibility choices.
Corley v. Jackson Police Department, 566 F.2d at 1000 n.8. This court has recently had occasion to comment on the need for adequate findings in order to provide a clear understanding of the basis of the decision. See H. Prang Trucking Co. v. Local Union No. 469, 613 F.2d 1235, 1239 (3d Cir. 1980); O’Neill v. United States, 411 F.2d 139, 146 (3d Cir. 1969) (“[I]t is necessary that the trial court specify these subordinate facts upon which the ultimate factual conclusion must rest.”)
Accordingly, we will remand this case for findings of fact as to whether there was comparability between Worthy’s safety rec*704ord and that of the white employees who he alleged had comparable safety records but who were not demoted or who committed comparable rule infractions but who were not subject to ten-day suspension.
In view of our decision to remand this case for additional findings, we deem it appropriate to consider some of the additional issues raised on appeal since they may arise in the course of the proceedings below. Appellant argues that the court erred in finding that those white employees who were demoted could be considered comparable to Worthy. Although the district court made the ultimate finding of comparability by its finding that Worthy’s safety record “placed him at the same level, if not worse, as white cranemen who were demoted”, it is not possible to ascertain from the opinion the factual basis for that ultimate finding. In the determination of comparability, the Supreme Court has cautioned that “precise equivalence” is not the ultimate issue. McDonald v. Santa Fe Trail Transportation Co., 427 U.S. at 283 n.11, 96 S.Ct. at 2580 n.11. Nevertheless, there must be some articulation of the circumstances to demonstrate that they are sufficiently parallel to permit the inference of comparability. In some cases, the comparability or lack of comparability may be self-evident. That is not the situation here where the safety violations are described variously as collisions, dropped block, smashed window of 80" speed pulpit with hoist block, damaged finishing mill speed pulpit, etc. Interpretation of these incidents is needed from the trial court who heard the testimony and examined the evidence in order that the ultimate finding of comparability can be reviewed.
Finally, appellant claims the fact that the employer demoted Worthy instead of assigning him to work as a coder in a position which did not require crane operation showed the pretextual nature of the employer’s actions. If appellant were correct in this claim, employers would be under an obligation to impose on the minority employee the minimum disciplinary sanction consonant with safety. The Supreme Court has explicitly stated that Title VII does not impose such an obligation on employers. In Furnco Construction Corp. v. Waters, 438 U.S. 567, 577-78, 98 S.Ct. 2943, 2950, 57 L.Ed.2d 957 (1978), the Court stated:
[The employer] need not prove that he pursued the course which would both enable him to achieve his own business goal and allow him to consider the most employment applications. Title VII prohibits him from having as a goal a work force selected by any proscribed discriminatory practice, but it does not impose a duty to adopt a hiring procedure that maximizes hiring of minority employees. . Courts are generally less competent than employers to restructure business practices, and unless mandated to do so by Congress they should not attempt it. (emphasis in original).
Accordingly we will remand the case for additional findings. Since we do not find that any evidence was improperly excluded, we believe there is no occasion for the trial court to adduce additional evidence.
For the foregoing reason's, the judgment of the district court will be vacated and the case will be remanded for further proceedings consistent with the opinion.
. Board of Trustees of Keene State College v. Sweeney, 439 U.S. 24, 99 S.Ct. 295, 58 L.Ed.2d 216 (1978) (per curiam); Furnco Construction Corp. v. Waters, 438 U.S. 567, 98 S.Ct. 2943, 57 L.Ed.2d 957 (1978); McDonnell Douglas Corp. v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973).