This article concerns the real-world importance of leadership for the success or failure of organizations and social institutions. The authors propose conceptualizing leadership and evaluating leaders in terms of the performance of the team or organization for which they are responsible. The authors next offer a taxonomy of the dependent variables used as criteria in leadership studies. A review of research using this taxonomy suggests that the vast empirical literature on leadership may tell us more about the success of individual managerial careers than the success of these people in leading groups, teams, and organizations. The authors then summarize the evidence showing that leaders do indeed affect the performance of organizations--for better or for worse--and conclude by describing the mechanisms through which they do so.