Adinazolam, a triazolobenzodiazepine, was compared with amitriptyline and placebo in a double-blind protocol involving melancholic depressives. Forty-eight patients entered the study, and 35 completed 6 weeks of evaluation. Patients were well matched in each of the groups. Initially, the adinazolam group showed more improvement than the placebo group, but this trend did not persist beyond day 7. The amitriptyline group tended to show more improvement than either the placebo or adinazolam groups, and final scores were significantly better for amitriptyline. Depressive symptoms in the amitriptyline dropout group appeared improved, whereas in the adinazolam and placebo dropout groups they were unchanged or worse. Both amitriptyline and adinazolam were more frequently associated with complaints of side effects than placebo. Thus, amitriptyline was observed to be superior to adinazolam and placebo in improvement in depressive symptoms while adinazolam had only a brief early advantage over placebo.