A wealth of data support a role for serotonin (5-HT) function in the mediation of satiety responses, that are impaired in patients with bulimia nervosa. Testmeal results are presented in which 26 bulimic patients and 17 normal controls were given in randomized, double-blind-fashion, placebo, and the 5-HT agents m-chlorophenylpiperazine (m-CPP, 0.5 mg/kg p.o.) and L-tryptophan (L-TRP, 100 mg/kg i.v.). Three and one-half hours after drug administration, subjects were allowed to eat and lib from a standardized testmeal of 3,500 calories, after which postprandial vomiting was not allowed. M-CPP, but not L-TRP, significantly decreased meal size in the combined group, the controls, and to a lesser extent, the bulimics (P < or = .06). Maximum m-CPP concentrations were inversely correlated to the number of calories consumed in the total group. Following m-CPP, there were significant decreases in carbohydrate, protein, and fat intake in the total group of subjects. There were also trends for decreased carbohydrate and protein intake in the bulimics following m-CPP. There were trends for both m-CPP and L-TRP to reduce fat intake in the controls. Differences in the effects between m-CPP and L-TRP are likely due to differential involvement of 5-HT receptor subtypes at presynaptic and postsynaptic sites. These studies in humans confirm reports in animals that m-CPP decreases food intake, including carbohydrates, protein, and fat in a mixed testmeal.