Safety aspects [adverse events, blood pressure and heart rate, weight, and laboratory tests (liver parameters, hemoglobin, leukocytes)] of long-term treatment in 1,120 patients are discussed. Adverse events during this long-term treatment were also compared with those of a subgroup of these patients who, before long-term treatment, were treated on a short-term basis (n = 706). Efficacy [Hamilton Depression Rating Scale (HAM-D), Clinical Global Impression of Efficacy (CGI), and occurrence of relapses and recurrences] in a homogeneous sample of 485 patients is also discussed. The adverse events most frequently observed during long-term treatment were insomnia, headache, and dizziness. Insomnia and headache were also most often occurring in the compared sample of patients with short-term treatment, whereas dizziness during this treatment period ranked at the fifth position. Supine and standing mean blood pressure did not consistently change during long-term treatment, the most prominent increases in comparison with baseline were seen in the period > 1 year of treatment (6.3 mm Hg supine/7.2 mm Hg standing). Comparison of blood pressure values in the hypertensive range at baseline and during long-term treatment revealed no statistical difference (McNemar test p = 0.07829). Mean heart rate slightly decreased during long-term treatment (by a maximum of 6.3 beats/min supine, 8.2 beats/min standing). Mean weight did not change between baseline and treatment end point. There were 23 patients with a weight loss of 10 kg or more and 16 patients with a weight gain of 10 kg or more. For none of the laboratory parameters tested was there a statistical significance regarding shifts from normal to pathological values. HAM-D mean total scores in the above-mentioned subgroup of patients decreased from 25.05 points at baseline (n = 485) to 7.88 points after 1 year of treatment (n = 139). Seventy-five patients who had favorably responded to treatment (total responders n = 300) relapsed during the first 6 months of treatment. During the second half-year of treatment the recurrence rate was 14.8%, and during the third 6 months the recurrence rate was 12.2%. CGI in the same subsample of patients in whom HAMD was evaluated (n = 485) compared with those patients who did not drop out during the short-term period up to 44 days and entered long-term treatment (n = 401) showed that the percentage of the ratings "no change/worse" was higher in the sample that also included patients who withdrew from treatment during the short-term period.(ABSTRACT TRUNCATED AT 400 WORDS)