The antidepressant efficacy and safety of venlafaxine was shown previously in 6-week, placebo-controlled trials. We evaluated the long-term safety and clinical acceptability of venlafaxine and imipramine in a double-blind, parallel-group, comparative study. Two hundred ninety depressed outpatients were treated with venlafaxine, and an additional 91 received imipramine for as long as clinically necessary, up to 1 year. The total daily dose of each drug could vary from 75 to 225 mg. The Clinical Global Impressions Scale and a therapeutic response rate that was based on Clinical Global Impressions Scale-Improvement and incorporated discontinuation information were used to evaluate efficacy. Safety determinations and patient subjective ratings were used to evaluate safety and clinical acceptability. During the study, the adverse events were generally mild to moderate and most subsided with continued treatment; the most frequent were nausea for venlafaxine and dry mouth for imipramine. The anticholinergic side effect burden was significantly higher in the imipramine group than in the venlafaxine group. Venlafaxine was judged significantly more acceptable than imipramine, on the basis of the subjective ratings by patients. Fewer venlafaxine-treated patients than imipramine-treated patients withdrew because of adverse events and unsatisfactory response. There was a consistent trend in the therapeutic response rates in favor of venlafaxine that reached statistical significance at months 2, 6, and 12. In this long-term study, patient acceptability was greater for venlafaxine than for imipramine, suggesting therapeutic advantages for venlafaxine in the long-term treatment of depression. Additional studies with other active comparators are underway to confirm and extend these encouraging results.