We review the controlled trials of antidepressant treatment in methadone patients. Several studies show antidepressant effects, but none demonstrate clear improvement in drug abuse. This is contrary to "self-medication" but rather suggests depression is either independent or substance induced. Methodologic limitations are noted, especially reliance on cross-sectional mood assessment, which may select transient mood disturbances rather than true affective disorder. We review our previously published pilot study of imipramine in depressed methadone patients selected by lifetime history, and we report four year treatment course in the nine patients who responded favorably during that trial. Patients remained euthymic during imipramine treatment and relapsed to depression during attempts to taper it. This suggests imipramine had an enduring antidepressant effect. However, intermittent drug use remained a problem for several patients, suggesting depression and drug abuse are at least in part independent disorders. Placebo controlled replications, combinations of antidepressant medication with psychosocial interventions, and exploration of antidepressants as adjuncts in methadone detoxification, are suggested avenues for further research.