Questionnaire data were collected from a panel of 342 gay men at risk for acquired immune deficiency syndrome enrolled in the Coping and Change Study between 1985 and 1987, and 1988 and 1990. Data were obtained across a period of 5 years in six serial wave pairs to determine the relationship of social support to Hopkins Symptom Checklist-assessed subsequent depression and general distress and to investigate whether the trends observed were stable or transient over time. Both objective and subjective components of social support demonstrated dramatic within-person stability over time (r = .47 to .86). A measure of subjective social support was modestly but significantly associated with lower depression at four of the six time periods and a lower level of general distress at one time period. Before appropriately controlling for current depression, subjective social support appeared to account for up to a third of the variance in future depression; after such controls were included in the regression equation, it became apparent that the independent contribution of support only ranged from 4% to 6% across the study period. This emphasizes the importance of including current mental health in longitudinal analyses. The respondents' social participation and involvement with others did not affect either depression or general distress at any time during the study period. These results indicate that while social participation may have no effect, subjective social support appears to influence often mental health in this cohort. Furthermore, human immunodeficiency virus seropositive men may at times benefit from such support.