A randomized, double-blind, parallel-group, 6-week study was undertaken to compare the efficacy and tolerability of once or twice daily administration of the selective serotonin reuptake inhibitors paroxetine and fluoxetine. After a 1-week placebo wash-out, patients suffering from DSM-III major depression and with a score of 18 or more on the 21-item Hamilton Rating Scale for Depression (HRSD) received either paroxetine or fluoxetine. The patients were assessed for efficacy using the HRSD, Montgomery-Asberg Depression Rating Scale and Clinical Global Impression; for tolerability, adverse events were elicited by the use of a non-leading question and a side effects checklist. The groups of patients were comparable on entry to the study. One hundred patients were recruited into the study, of whom 78 were evaluable for the efficacy analysis. Paroxetine and fluoxetine showed comparable efficacy at the end of the 6-week treatment period, but a statistically significant difference in the number of responders at week 3 in favour of paroxetine was observed. This could suggest an earlier onset of action with paroxetine. Also, associated anxiety symptoms were significantly reduced on paroxetine compared with fluoxetine at week 3. Patients on paroxetine reported fewer adverse events than those on fluoxetine. The most commonly reported adverse events were nausea and vomiting in both groups.