We evaluated treatment response for depressed versus nondepressed ambulatory cocaine abusers in a 12-week randomized controlled trial of desipramine and cognitive-behavioral treatment, alone and in combination. Subjects with depressive symptomatology at baseline tended to have better retention and better cocaine outcomes compared with nondepressed subjects. Desipramine was an effective antidepressant in this sample and was associated with significantly greater reduction in depressive symptoms than was placebo; however, desipramine treatment was not associated with greater reductions in cocaine use for either the depressed or euthymic subgroup. Cognitive-behavioral relapse prevention treatment was associated with significantly longer periods of consecutive abstinence and better retention compared with supportive clinical management for the depressed subgroup, but psychotherapy condition did not have an effect on depressive symptoms. These data point to differential symptom reduction in depressed cocaine addicts and underscore the importance of evaluating combined psychotherapy-pharmacotherapy approaches for this population.