One pitfall of organizational diversity, especially in superior-subordinate dyads, is that rater-ratee dissimilarity is negatively associated with scores on performance appraisals. The rater-ratee dissimilarity can stem from demographics (e.g., race, gender, age), attitudes (e.g., conservatives and liberals), personality characteristics (e.g., conscientiousness), or other factors, such as group membership. Social identity theory and self-categorization theory can be used to explain rating bias. In short, these theories predict that persons classify the self and others into groups. Persons similar to the self are considered ingroup members while persons dissimilar to the self are considered outgroup members. This categorization results in an intergroup bias, whereby persons have more positive evaluations of ingroup members than they do outgroup members. Therefore, if the categorization process is at the root of the bias in performance appraisals, then techniques targeted at transforming the categorization boundaries could mitigate these effects. The current study examined this possibility in examining the impact of recategorization (i.e., the process that enables persons to conceive of themselves as belonging to a single, common group) on performance appraisals. Data were gathered from 66 participants in a posttest only control group design experiment. Participants first worked in 3-person groups to establish categorization boundaries. Cooperation (i.e., interaction, common problem, and common fate) was used to recategorize members. The manipulation was successful, as persons in the cooperative condition were more likely to perceive the aggregate to represent one group than were those in the no-cooperation condition (93% vs 36%). An analysis of covariance indicated that participants who perceived the aggregate to represent one group had less intergroup bias toward outgroup members, F (1, 56)=3.21, p < .10. In turn, hierarchical regressions indicated that intergroup bias was inversely related to performance appraisals. Finally, tests for mediation (Baron & Kenny, 1986) revealed that intergroup bias partially mediated the relationship between perceptions of the aggregate as one group and intergroup bias. Results demonstrate the efficacy of recategorization in managing diversity, especially with respect to performance appraisals.