In a 2-week study, 24 newly admitted manic patients were treated for 1 week with L-tryptophan (12 g/day); during the second week, half the patients, chosen at random, continued to receive tryptophan, while placebo was substituted in the other half under double-blind conditions. In the open phase of the study, there was a clinically and statistically (p less than 0.001) significant reduction in manic symptom scores, with little need for haloperidol prn. Patients who continued to be treated with tryptophan showed no significant change in mean scores during the second week, but those who were switched to placebo tended (p less than 0.10) to show an increase in the mean scores for manic symptoms. There was a significant (p less than 0.05) increase in the geometric mean of morning fasting total and free plasma tryptophan concentrations in men, but not in women. These results suggest that increasing the synthesis of 5-hydroxytryptamine has some therapeutic effect in mania.