Brain catecholamines can be rapidly reduced by inhibiting their synthesis with alpha-methyl-para-tyrosine (AMPT). In order to assess the role of catecholamines in antidepressant action AMPT challenges were administered in a double-blind, placebo-controlled, crossover fashion to 14 depressed patients having maintained a therapeutic antidepressant response for > or = 2 weeks (3 desipramine, 2 mazindol, 5 fluoxetine, 4 sertraline). Each patient participated in two challenges one week apart. Each challenge included a baseline, two days of either AMPT or diphenhydramine (active placebo), and a followup. Antidepressant drugs were continued throughout testing. The 3 desipramine- and 2 mazindol-responders had a rapid increase in depression score during AMPT but not placebo (diphenhydramine) challenge whereas only 1 of 9 selective serotonin reuptake inhibitor (SSRI)-treated patients did. The implication of this is that the antidepressant response to desipramine may be more acutely dependent on brain catecholamine content than the response to SSRIs. In the context of our previous work with tryptophan depletion, these results suggest that the neurobiological mechanisms underlying the antidepressant responses to different drugs involve alterations in the functioning of different neurotransmitter systems and reinforce the importance of changes in both the serotonin and catecholamine systems for successful antidepressant responses.