21 depressed patients of the Basle University Psychiatric Outpatient Clinic were treated in a double-blind study with paroxetine and amitriptyline. 11 of these patients did not continue the trial until the end of the 7th week. There was a significant difference in the number of dropouts between the two groups: 80% of the amitriptyline group did not continue until the end, while in the paroxetine group we found only 30% dropouts. The patients of both groups showed a gradual decrease of the median total scores on the Hamilton and the Montgomery and Asberg Depression Rating Scales. Although the number of patients who stayed in the trial for at least 4 weeks (8 with paroxetine, 6 with amitriptyline) is quite small, we see from the results of the clinical global impression that the members of the paroxetine group improved most of all in the somatic symptoms, while considering their moods we found no differences between the groups. Patients of both groups complained about side effects, most of all about dry mouth and tiredness. From the high rate of dropouts under amitriptyline we found that the side effects under this drug were more severe and therefore led to the dropouts.