This study examined whether a memory or response bias mediated the effects of performance cues on observers' recollections of a work group's behavior. Fifty-nine subjects observed a film of a group at work. Then, immediately or one week later, subjects rated the group's performance using behavioral, evaluative, and objective outcome rating instruments. Prior to observing the group, subjects were given performance cues that led them to believe the group had either performed well or poorly in the task. Results identified a systematic response bias and not a memory bias as the cognitive process mediating the effects of performance cues in work behavior ratings: Subjects adopted a more liberal decision criterion when attributing effective (ineffective) behaviors in response to positive (negative) performance cues. Unlike the behavioral ratings, recollections of specific group outcomes were immune to the biasing effect of performance cues.