The relation between level of physical activity and risk of subsequent depression was examined using three waves of data from the Alameda County Study. Among subjects who were not depressed at baseline, those who reported a low activity level were at significantly greater risk for depression at the 1974 follow-up than were those who reported high levels of activity at baseline. Adjustments for physical health, socioeconomic status, life events, social supports, and other health habits did not affect the association appreciably. Associations between 1965-1974 changes in activity level and depression in the 1983 follow-up suggest that the risk of depression can be altered by changes in exercise habits, although these associations were not statistically significant after adjustment for covariates. These results provide somewhat stronger evidence for an activity-depression link than do previous studies, and they argue for the inclusion of exercise programs as part of community mental health programs, as well as for further studies that focus on the relation between life-style and mental health.