The aim of this multicenter, randomized, double-blind, 8-week study was to compare the antidepressant efficacy and tolerability of mirtazapine and venlafaxine in the treatment of hospitalized patients with DSM-IV diagnosis of severe depressive episode with melancholic features. Patients with a baseline score of > or = 25 on the 17-item Hamilton Rating Scale for Depression (HAM-D-17) were randomly assigned to receive treatment with either mirtazapine (N = 78, 15-60 mg/day) or venlafaxine (N = 79, 75-375 mg/day, twice a day) in a rapid up-titration schedule. Efficacy was assessed with the Montgomery-Asberg Depression Rating Scale (MADRS), HAM-D-17, and Clinical Global Impression scale, and quality of life was assessed with the Quality of Life, Enjoyment, and Satisfaction Questionnaire and Quality of Life in Depression Scale. Tolerability was assessed with the Utvalg for Kliniske Undersogelser (UKU) side effect scale and by reporting adverse events. Both drugs were effective in reducing overall symptoms of depression, showing substantial reductions in group mean MADRS scores (-20.1 for mirtazapine and -17.5 for venlafaxine) and HAM-D-17 scores (-17.1 for mirtazapine and -14.6 for venlafaxine) at the end of the treatment. Although not statistically significant, at all assessment times higher percentages of patients treated with mirtazapine were classified as responders (> or =50% reduction) on the HAM-D (at endpoint, 62% vs. 52%) and MADRS (at endpoint: 64% vs. 58%). Likewise were the percentages of remitters (HAM-D score < or =7; MADRS score < or =12) also higher in the mirtazapine group. A statistically significant difference favoring mirtazapine was found on the HAM-D Sleep Disturbance factor at all assessment points (p < or = 0.03). Both treatments were well tolerated. Although slightly more subjects treated with mirtazapine reported at least one adverse event, a statistically significantly higher percentage of patients treated with venlafaxine (15.3%) than mirtazapine (5.1%) dropped out because of adverse events (p = 0.037). Quality of life improved in both treatment groups. In this study, treatment with mirtazapine resulted in a trend toward more responders and remitters than treatment with venlafaxine and in significantly fewer dropouts as a result of adverse events.