Three studies were conducted to examine the predictions that (a) in-group identification depends on optimal distinctiveness needs (Study 1), and (b) that social identity threat overrides the predictive value of these needs to determine identification (Studies 2 and 3). In Study 1, need for assimilation and need for differentiation were assessed among natural groups. We found support for the optimal distinctiveness theory (ODT) prediction that there is a curvilinear relationship between identification and optimal distinctiveness needs satisfaction. In Studies 2 and 3, interactive effects of the extent to which groups satisfy assimilation and differentiation needs (groups are either too small, of moderate size, or too large) and social identity threat were examined. In the no identity threat condition identification was higher in moderately sized groups (where both needs are balances) compared to groups were either assimilation or differentiation dominates. However, when facing an identity threat, identification was highest in very small groups, providing evidence that social identity concerns override individual need satisfaction. Discussion focuses on comparing and integrating ODT and social identity theory.