The efficacy and tolerability of the selective reversible monoamine oxidase A inhibitor, moclobemide (300 mg/day) and the selective serotonin uptake inhibitor, fluoxetine (200 mg/day), were compared in a six-week single-centre double-blind fixed-dose study in patients (n = 42) with double depression (DSM-III-R: dysthymia with superimposed major depressive episode) using weekly assessment on the Hamilton depression rating scale (HDRS-17 items) and clinical global impression (CGI) scale. The primary efficacy outcome measure was a decrease > or = 50% in end of treatment HDRS score, secondary measures were the mean total endpoint HDRS scores and percentages of CGI very good and good responses. Tolerability was measured by the frequency and severity of volunteered adverse events. There were no significant differences in secondary efficacy outcome measures, but more patients achieved a > or = 50% decrease in HDRS score on moclobemide (71% vs 38%, p < 0.05). The only adverse event was mild transient anxiety (n = 1) with moclobemide. The results suggest that moclobemide and fluoxetine are equally well tolerated and at least similar in efficacy in double depression. Evidence that moclobemide may be more effective requires confirmation in a larger comparative study incorporating a placebo control group.