BACKGROUND: This study was undertaken to compare the antidepressant efficacy and short-term safety of venlafaxine with those of placebo in hospitalized patients with major depression and melancholia. METHOD: Ninety-three inpatients with a minimum prestudy Montgomery-Asberg Depression Rating Scale (MADRS) score of 25 were treated for up to 4 weeks with either venlafaxine or placebo. Dosage averaged approximately 350 mg/day during Weeks 2 to 4. Efficacy and safety were assessed throughout the study. Efficacy was evaluated using the MADRS, the 21-item Hamilton Rating Scale for Depression (HAM-D), and the Clinical Global Impressions (CGI) scale. Recorded study events, vital signs and body weight measurements, laboratory determinations, physical examinations, and ECG recordings were used to assess safety. RESULTS: Venlafaxine provided significantly greater improvement in the MADRS scores after 4 days and in the HAM-D scores after 1 week than did placebo. Response rate (based on a 50% decrease in MADRS scores) was 65% (30 of 46 patients) for venlafaxine and 28% (13 of 47 patients) for placebo. Significantly more placebo-treated patients (40%; N = 19) than venlafaxine-treated patients (9%; N = 4) discontinued treatment early because of lack of efficacy. Nausea and sweating were the most common events, occurring at a significantly higher rate in the venlafaxine group. CONCLUSION: Venlafaxine is an effective and well-tolerated antidepressant in hospitalized patients with major depression and melancholia.