We examined the efficacy and safety of three different dosages of venlafaxine hydrochloride (75, 225, and 375 mg/day) in a multicenter, randomized, double-blind, placebo-controlled, four-group study. Outpatients, 18 to 65 years old, who met DSM-III criteria for major depression were included (N = 358 randomized; 194 completed). Of the total patients completing the trial, 59%, 56%, 51%, and 51% were in the placebo, 75-mg, 225-mg, and 375-mg groups, respectively. The primary outcome measures were the Hamilton Rating Scale for Depression (HAM-D21) total, HAM-D21 depression item, Montgomery-Asberg Depression Rating Scale total, and Clinical Global Impressions scale. Each dosage of venlafaxine was associated with statistically significant improvement as compared with placebo, based on the intent-to-treat sample. The two higher dosages were associated with a modestly greater antidepressant response than was the 75-mg dosage. Nausea, dizziness, somnolence, and anorexia were the most common adverse events attributable to venlafaxine. Since headache occurred at a similar frequency in both the drug and placebo groups, we did not consider it to be attributable to venlafaxine use. Withdrawal from the study due to adverse events occurred in 5%, 17%, 24%, and 30% of the patients in the placebo, 75-mg, 225-mg, and 375-mg groups, respectively. Venlafaxine, at dosages of 75-375 mg/day, is an effective and well-tolerated antidepressant. With increasing dosage, greater efficacy and possibly more adverse effects will occur.