The present experiment provided a test of Fiedler's (1967) Contingency Model of Leadership Effectiveness, i.e., the relationship of leader style to group effectiveness is mediated by situational demands. Thirty-two 4 man task groups composed of military academy cadets were run in the experiment. In accordance with the Contingency Model, leaders were selected and assigned to groups on the basis of a measure of leadership style. Groups were assigned to one of 8 task situations specified by the manipulation of the variables of leader acceptance, task structure, and leader power. Measures were made of group task productivity, group atmosphere, and leader behavior. Results on measures of group productivity give extremely strong support for the validity of the Contingency Model. Measures of leader behavior help to illuminate the processes which result in the differential effectiveness of varying leadership styles across situations. (Author)