Pharmacotherapy response in depressed (n = 20) vs. nondepressed (n = 74) cocaine-abusing methadone maintenance patients was compared in a 12-week, randomized, double-blind trial using amantadine at 300 mg daily (n = 33), desipramine at 150 mg daily (n = 30), and placebo (n = 31). Starting in Week 3, the depressed patients on medications reported significantly less cocaine usage than the depressed patients who received placebo. By Week 10, the medicated depressed patients (in treatment) reported a 96 percent decrease in cocaine usage and a 68 percent decrease in cocaine craving. During the last 2 weeks of treatment, 42 percent of the urine toxicologies of the medicated depressed patients were cocaine free, compared with only 6 percent of the placebo depressed patients. Also, the placebo nondepressed patients had significantly better treatment outcome compared with the placebo depressed patients. Thus, depression appears to be an important predictor of poor treatment outcome with relapse prevention therapy alone and of good response to cocaine abuse treatment with medication.