The purpose of this article was to determine proximal and distal antecedents of leadership in the U.S. Army's Officer Candidate School (OCS). A model composed of motivation to lead, leadership self-efficacy, implicit leadership, organizational commitment, general cognitive ability, and personality was proposed. Results from a longitudinal examination of 1,232 officer candidates suggest partial support for the model, and limited evidence for enlisted experience as a moderator. Candidate personality (partially mediated by interest in leadership and leadership self-efficacy) best predicted leadership performance during OCS and peer ratings of leadership potential. Implications for OCS selection and models of leadership performance are discussed.