Patients (n = 490) suffering from a major depressive episode according to DSM-III criteria were randomly allocated to groups receiving either moclobemide, imipramine, or placebo treatment. Subjects were treated as out-patients for 6 weeks. On overall assessment of efficacy and on results of the Hamilton Rating Scale for Depression, both moclobemide and imipramine were superior to placebo, but the differences between moclobemide and imipramine were not significant. Premature termination due to insufficient efficacy was more frequent with placebo than with moclobemide or with imipramine, these differences being significant. The overall assessment of tolerance clearly favoured placebo and moclobemide over imipramine. This was also reflected in the frequency of premature terminations due to poor tolerance, as well as in the frequency of adverse events, which were highest in the imipramine group. The only cardiovascular finding was an increase of the mean heart rate with imipramine, maximum at the end of week 1, while placebo and moclobemide displayed no relevant changes. There were no other important drug-related changes.