Central (intracerebroventricular) and peripheral (intraperitoneal) injections of the octapeptide of cholecystokinin (CCK-8) were compared to determine the most effective route of administration to elicit satiety for food intake in the rat. Subdiaphragmatic bilateral vagotomy and spinal cordotomy (T2-T3) were also performed to investigate the importance of visceral nerves for the satiety effect. CCK-8 suppressed feeding and elicited satiety resting behavior when injected peripherally but it was less effective when injected centrally. The satiety effect of CCK-8 or CCK-33 following peripheral injections was blocked by vagotomy whereas spinal cordotomy had no effect. The results indicate that some component of the vagus is required to mediate the peripherally induced cholecystokinin satiety effect, but the splanchnic nerves are not necessary. The weak effect of CCK-8 following ventricular administration is additional evidence suggesting that cholecystokinin of intestinal origin acts in the periphery rather than directly on the brain to elicit its typically rapid satiety effect in rats.