Twelve nondepressed healthy female obese subjects (BMI > 30 kg/m2) took part in a study which conformed to a double-blind randomized crossover design. Each subject acted as her own control across 2 weeks of treatment with either 60 mg of the 5-HT reuptake inhibitor fluoxetine or matching placebo. On days 7 and 14 of both treatment phases subjects were provided with fixed energy lunch meals high in either CHO or fat. The effect of these meals on satiety during the fluoxetine and placebo phases was assessed by a battery of procedures. Subjects felt less hungry after consuming the high CHO meal than after consuming the high-fat meal. They also felt less hungry when taking fluoxetine than when taking the placebo. Analysis of energy intake from the test meal revealed a main effect of prior lunch meal type (high CHO or high fat) and a main effect of drug treatment. Subjects consumed an average of 574 kcal following the high CHO meal compared to 689 kcal following the high-fat meal. Subjects also consumed an average of 532 kcal when taking fluoxetine compared to 730 kcal when taking the placebo. Fluoxetine did not exert any significant effects on macronutrient selection. Mean daily energy intake, calculated from food diary records, was 1881 kcal when subjects were taking the placebo compared to 1460 kcal when taking fluoxetine (a reduction of 22.4%). Fluoxetine treatment produced a significant weight loss of 1.97 kg over the two weeks of treatment compared to a weight loss of only 0.04 kg on placebo.