The controversy about the causal ordering of leader behaviors and unit performance is reviewed. The analysis indicates that neither the behavior-performance nor the performance-behavior proponents have provided convincing evidence. A longitudinal study of the development of these interrelationships from the initial stages is described. The results suggest that group members perceive a behavior-performance ordering but leaders perceive a performance-behavior relationship. These results imply that further research should focus upon both parties to the relationship in the earliest phases of development. Since both parties appear to make decisions with limited information, an attributional perspective may be needed. (Author)