The authors reviewed the efficacy and safety of stimulant drugs in the treatment of depression. Although uncontrolled studies were generally positive, the 10 placebo-controlled studies of stimulant drugs in primary depression, with one exception, indicated little advantage of drug over placebo. Although several of these studies were methodologically unsophisticated, they were comparable with and performed during the same period as studies establishing the efficacy of imipramine. Controlled studies of stimulants in apathetic or depressed geriatric patients were more likely to be positive, but outcome frequently consisted of partial improvement. Studies in medically ill patients with depression were promising but uncontrolled. Side effects have not been severe, and these drugs may pose less of a risk than tricyclics in the medically ill or elderly. Habituation is suggested, but there are no placebo-controlled studies to confirm this. In short, the stimulant drugs do not appear to be as effective as the conventional antidepressants in primary depression, but may be of value in refractory cases or in special cases, such as those involving the medically ill patient. Placebo-controlled trials are needed to explore these questions.