Fluoxetine, a selective inhibitor of 5-HT uptake, was compared to dothiepin in a double-blind study of 6 weeks duration in 100 depressed patients (male and female) drawn from 8 general practices. Only those who scored at least 17 on the first 17 questions of the Hamilton Psychiatric Rating Scale for Depression (HAM-D) were selected. Both groups improved throughout the trial, though the dothiepin treated patients tended to improve quicker. However, by the end of the trial there was no statistically significant difference between the 2 groups. Subset analyses of HAM-D scores associated with anxiety and sleep revealed no statistically significant differences between the 2 treatments though improvement in anxiety scores was marginally greater for those receiving fluoxetine by the end of the trial. Other global assessments by patients and doctors confirmed the changes in HAM-D scores. Statistically significant weight changes occurred between visits 1 and 5. Whereas fluoxetine-treated patients lost weight (p less than 0.05), dothiepin-treated patients gained weight (p = 0.05) over this period. Adverse effects were reported in 27 patients given fluoxetine and 20 dothiepin. Of these, 14 fluoxetine and 7 dothiepin-treated patients withdrew before the end of the trial. The most common adverse effects were nausea, vomiting and diarrhoea in the fluoxetine group and tiredness, drowsiness and diarrhoea in the dothiepin group. There were no haematological or clinical chemistry changes.