ABSTRACT Introduction Traits, including leader identity, represent cross‐situational consistency in responses. Cross‐situational variability is treated as random error. We explore cross‐situational variability in leader identity goals as a function of self‐evaluations and construct an individual difference variable, contingent leader identity, which we hypothesis will supplement the effects of trait leader identity in prediction of leader emergence. Methods Data were collected biweekly over 36 weeks from 263 participants in 35 single‐sex, initially leaderless student groups. Leader emergence ratings were provided by peers. Multilevel analyses using Mplus 8.6, incorporated within‐person ratings of self‐evaluations of leadership ability and leader identity, peer ratings of leadership, and control variables. Participants were nested within teams and completed up to 18 questionnaires, creating a three‐level design. Slope betas from regression of leader identity on self‐evaluations provided the measure of individual contingent leader identity. Results Contingent leader identity scores ranged from −0.48 to 0.84 and were stable across the first and second half of the study ( r = 0.75, p < 0.001). As predicted, individual's contingent leader identity was a significant predictor of their recognition as a leader by their peers in final weeks of the study. Conclusion Contingent leader identity and related CAPS constructs provide an explanatory mechanism for leader behavior in descriptive models of leadership. They also offer the prospect of retesting contingency theories of leadership to establish if leaders do vary their behavior across situations and if situation‐behavior contingencies predict effectiveness as proposed in contingency models. Leadership development programs should focus on self‐evaluation skills, not just leadership behaviors.