In a double-blind clinical trial comprising 29 depressed patients citalopram, a highly selective 5-HT re-uptake inhibitor and maprotiline, a specific NA re-uptake inhibitor, were compared. Allowing for the small sample and taking into consideration that both groups consisted of severely ill, hospitalized patients, it is notable that half of them appeared to respond to treatment. Comparison of the clinical efficacy of the two drugs showed no significant difference, but the profiles of the side-effects appeared to be different. The patients treated with citalopram showed increased sweating, drowsiness, restlessness and headache. These side-effects were almost entirely reported by the non-responders. The maprotiline patients had anticholinergic symptoms, such as dryness of mouth and constipation, side-effects which were also reported by the responders. No correlation was found between plasma steady-state levels of either drug and clinical outcome. The Dexamethasone Suppression Test (DST) appeared to show some predictive value as regards treatment response. There was a tendency towards better overall treatment results in the non-suppressor group. Determination of post-probenecid 5-HIAA, HVA and MHPG concentrations in lumbar-CSF was made in 22 patients. There was a significant negative correlation between HVA and the severity of depression, as well as a significant negative correlation of MHPG with the Newcastle score. The 5-HIAA concentration was found to be correlated with HVA, but not with MHPG. Rather surprisingly significant negative correlation between 5-HIAA and treatment results with maprotiline was found, but no correlation with MHPG. The lumbar-CSF MHPG and HVA values did not appear to have any predictive value as regards treatment response to citalopram or maprotiline. As expected the serotonin (5-HT) concentration in blood and thrombocytes in patients treated with citalopram showed a highly significant reduction after 2 and 4 weeks of treatment.