To investigate the effects of the long-acting opiate antagonist naltrexone on spontaneous human eating behavior, eight moderately obese male paid volunteers were housed in a hospital metabolic unit for 28 days and offered palatable foods ad lib by a platter service method. Under double-blind conditions, equally divided doses of 100, 200 and 300 mg naltrexone, or an acetaminophen placebo, were administered twice daily in tablet form for 3-day periods each, according to a Latin Square design. The doses of naltrexone resulted in decreases of daily caloric intake from placebo level, but these reductions were neither statistically significant nor dose-related. When the averaged effects of the doses were compared to placebo, five subjects showed intake reductions but the overall intake reduction of 301.5 +/- 198.1 kcal/day (mean +/- SEM) was not statistically significant. Naltrexone administration failed to selectively alter intakes of individual meals and snacks or macronutrient consumption patterns. During active drug periods, subjects lost 0.62 +/- 0.22 lb over 3 days, while during the placebo period, subjects gained 0.46 +/- 0.68 lb. However, there was no reliable change of basal metabolic rate as a function of naltrexone administration. The present results, which indicate that naltrexone administration is relatively ineffective in reducing food intake and inducing body weight loss in obese humans, are thus in contrast with reports that administration of opiate antagonist agents promote significant reductions of food intake and attenuations of body weight gain in experimental animals.