Hypotheses generated from animal studies that the endogenous opioid system is an important modulator of food intake suggest that blockade of the system in humans should affect eating behavior. To assess this hypothesis, seven normal volunteers were given 2 mg/kg naloxone or placebo on separate days in a double-blind, random but balanced cross-over experimental design. Compared to placebo, naloxone was found to reduce significantly total food intake from preselected prepared trays served 2.75 and 7.75 hours after drug administration (p < 0.02). The reduction was considerable (28%), and although the magnitude varied greatly among individuals, reduction occurred in each. This reduced food intake was not accompanied by a demonstrable alteration of the volunteers' perceptions of their hunger. Further cautious experimental investigation of naloxone's effects during long-term administration and in patients with eating disorders is warranted in light of its apparent effect of reducing food intake in humans while not decreasing their satiety.There is substantial evidence from animal studies that the endogenous opioid system (EOS) is an important modulator of eating behavior (1, 2). An important component of this evidence is the doserelated suppression of food intake produced by the administration of naloxone in the mg/kg range to rodents (3, 4). Because naloxone is considered a pure opiate receptor antagonist, this effect is consid- ered to reflect functional blockade of the EOS and thus active involvement of the EOS in the eating behavior of these animals.The evaluation of behavioral effects produced by the administration of naloxone has also been a major experimental approach in clinical research. However, clinical research began with the use of doses as low as 5 (xg/kg (0.4 mg) and still has not generally exceeded 0.3 mg/kg (20 mg) (5-7). These doses originated from the exceedingly small dose (less than 5 |xg/kg) needed to precipitate withdrawal from narcotic alkaloids (exogenous opioids) in humans (8).We have recently demonstrated dosedependent behavioral effects of naloxone given to normal volunteers in the mg/kg range that were not evident in previous clinical studies using lower doses (9, 10). Concomitant physiologic and hormonal evaluations supported the continuing specificity of naloxone as an opiate receptor antagonist at these doses. These data suggested that some EOS were not completely blocked at lower doses of naloxone, supporting the presence in humans 132