OBJECTIVE: Authors studied the efficacy and tolerability of mirtazapine and paroxetine in elderly patients with major depression during an acute phase (8 weeks) and an extension phase (16 weeks). METHODS: Patients with major depression and without dementia, at least 65 years old, were eligible; they were randomized to mirtazapine or paroxetine once daily, with doses increasing over 42 days. Efficacy was assessed with the Ham-D and Clinical Global Impressions Scale, and tolerability was assessed from adverse events. RESULTS: Of 255 patients randomized, 126 on mirtazapine and 120 on paroxetine were included in the efficacy analysis. Differences favoring mirtazapine were observed for the mean change from baseline in Ham-D-17 score. Other significant differences were in the proportion of patients classified as responders (50% decrease from baseline Ham-D-17 scores) at Day 14 and in remission (Ham-D-17 score of 7 or less) at Day 42. The median time to response was 26 days in the mirtazapine group and 40 days in the paroxetine group. The mirtazapine group also showed more reduction in Ham-D Factor I (Anxiety/Somatization) and Factor VI (Sleep Disturbance) scores. Efficacy of both drugs was maintained during the extension phase. Patients on paroxetine were more likely to discontinue therapy in the acute phase because of adverse events. CONCLUSION: During the first weeks of treatment, antidepressant effects were more pronounced in the mirtazapine group, suggesting that mirtazapine has an earlier onset of action. Mirtazapine also demonstrated a better tolerability profile and represents a valuable option for the treatment of depression in elderly patients.