A previously tested theoretical model that specifies relationships among depression, activities of daily living (ADL) impairment, and social support components was validated with 307 patients (aged 60 and older) from a multi-site effectiveness trial of 3 treatments (antidepressant, placebo, Problem-Solving Treatment) for dysthymia or minor depression in primary care. Ss completed interviews and self-reports at baseline and at 6 and 12 wks. The short-term, longitudinal data were analyzed with covariance structure modeling techniques. Consistent with the previous model, impairment in ADLs was associated with subsequent increases in depression; a larger emotionally close network that made frequent visits was associated with subsequent increases in perceived support; and perceived support was associated with subsequent decreases in depression. This last effect was significant only among participants randomly assigned to receive placebo with clinical management. The effect of perceived support primarily in the placebo group suggests that those with greater perceived social support and subsyndromal depression may be more likely to have a positive response to nonspecific clinical treatment components.