The research team theorizes that social group members’ numerical representation in an organization, compared with the majority group, influences concerns about their distinctiveness, and consequently, whether diversity approaches are effective. They combine laboratory and field methods to evaluate this theory in a professional setting, in which White women are moderately represented and Black individuals are represented in very small numbers. The team demonstrates that Black individuals report greater representation-based concerns than White women (Study 1). Next, they observe that tailoring diversity approaches to these concerns yields greater performance and persistence (Studies 2 and 3). The researchers then manipulated social groups’ perceived representation and find that highlighting differences (vs. equality) is more effective when groups’ representation is moderate, but less effective when groups’ representation is very low (Study 4). Finally, the team content-codes the diversity statements of 151 major U.S. law firms and find that firms that emphasize differences have lower attrition rates among White women, whereas firms that emphasize equality have lower attrition rates among racial minorities (Study 5).