We followed a cohort of subjects (predominantly inpatients) suffering a major depressive episode in midlife and late life for 1 year (N= 118). In this follow-up study, we examined three hypotheses, a) Elder subjects suffering major depression, compared with middle-aged subjects, will more likely report endogenous symptoms and less likely report decreased life satisfaction symptoms at 1-year follow-up. b) Impaired social support during the index episode will predict poor life satisfaction, but not endogenous symptoms at 1-year follow-up, regardless of age. c) Impaired social support during the index episode will be more predictive of decreased life satisfaction symptoms in midlife, compared with late life, at 12-month follow-up. The first two hypotheses were not supported, but the third hypothesis was. Both decreased life satisfaction and endogenous symptoms at outcome were significantly predicted by impaired social support during the index episode. Impaired social support predicted a poor outcome from an episode of major depression in both middle life and late life in controlled analyses. However, the effect of impaired subjective social support was conditional on age. Subjective social support appears to have a decreasing influence on the report of both endogenous and decreased life satisfaction symptoms for older individuals.