The extent to which subjects reciprocate unfavorable evaluations of performance was investigated in a dyadic inspection situation. The members of twelve male subject pairings alternated in their roles as ‘inspector’ and ‘operator’ on a discrimination task. As operators, the subjects were led to believe that their inspectors had evaluated their performance favorably or unfavorably on the basis of either a subjective judgment or an objective matching evaluation criterion. Results showed that those operators who reciprocated most were those whose discriminations were rejected by means of the judgment criterion. This result is in line with the attribution hypothesis that a person is held more responsible for an act having an unfavorable outcome for others when that act is perceived as voluntary and intended than when it is perceived as compulsory or externally determined.