Results on the use of sustained high doses of levodopa in depressed patients are presented, and the psychiatric side effects encountered during the therapeutic use of the drug in Parkinson's disease are discussed against the background of the hypothesized role of catecholamines in the pathogenesis of affective disorders. One half of the patients were treated with smaller doses of levodopa in combination with a peripheral decarboxylase inhibitor (MK‐485). Only 6 of 21 depressed patients consistently improved on levodopa, with relapse following placebo substitution and remission on restoration of the drug. A statistically significant increase in anger rating following 5 days of maximal dose of levodopa was noted as compared with that prior to treatment; others who previously experienced manic episodes manifested hypomanic behavior during treatment. Doses large enough to produce hypomania did not reverse depression. It is concluded that: (1) Amine depletion does not explain the pathophysiology of depression in most patients. (2) Hypomania occurs almost exclusively in affectively labile individuals. (3) The catecholamine hypothesis for depression may be an oversimplification in that mania and depression do not appear to represent the opposite poles of a single biochemical continuum. (4) Many factors contribute to psychiatric side effects—the pre‐dopa psychiatric state being the most important variable, levodopa often serving to unmask pre‐existing psychopathology.