A structural equation analysis of data collected from 484 members of the Emergency Medicine Residents Association was used to test a model in which peer support and work-group cohesiveness predicted role ambiguity. Role ambiguity was then specified as a predictor of work-related stress. Stress was hypothesized to affect depression and work satisfaction. The analysis confirmed the direct relationship between stress and depression and indicated that support from peers and the work group reduces stress. This impact is influenced by the amount of role ambiguity perceived by residents. Peer support, role ambiguity, and stress explain 52% of the variance in depression reported by residents. Residents are more satisfied with their work when their peers are supportive and when levels of occupational stress and role ambiguity are not high. The model accounted for 47% of the variance in reported work satisfaction. The results suggest that role ambiguity leads to perceptions of stress. This perceived stress, in the absence of strong support groups, results in increased depressive symptomatology and decreased work satisfaction.