In a replication and extension of the study by M. E. Heilman, M. C. Simon, and D. P. Repper (1987), 201 undergraduates participated in a simulation in which they experienced differing selection procedures and outcome feedback. Selection procedures did not have the deleterious effects on women that were found previously. Instead, race interacted with gender to moderate this relationship, and outcome played a significant role in participant self-evaluations. Black participants rated their leadership ability highest when both chosen preferentially and given negative outcome feedback, apparently because of a desire to maintain positive self-esteem. The latter explanation was supported in a follow-up study in which undergraduates (n = 80) worked in groups and received negative outcome feedback from either a racially similar or racially different experimenter. Theoretical and practical implications relating to diversity and self-appraisal management are discussed.