The slowing of gastric emptying by hyperosmolar solutions has been postulated to result from the triggering of duodenal osmoreceptor feedback on the stomach. We tested the idea that the inhibition of gastric emptying by a hyperosmolar solution depended on the duodenal resistance and the triggering of nutrient-specific feedback by tracking gastric emptying of 300 and 1,200 mosmol/kgH2O test solutions in 12 dogs in which duodenal resistance was either removed (by temporarily diverting chyme from uncorked duodenal fistula) or preserved (by keeping duodenal fistula corked). Mannitol was used to test osmolality alone, and glucose was used to examine the combined effects of osmolality and nutrient-specific inhibitory feedback. We found that: 1) the slowing effect of hyperosmolality was more marked with the duodenal resistance preserved (P < 0.05; analysis of variance), 2) the slowing effect of glucose was greater than that of mannitol for all conditions (P = 0.01; analysis of variance), and 3) the inhibitory effect of mannitol was localized to the duodenum. We conclude that inhibition of gastric emptying by hyperosmolar mannitol depended primarily on duodenal resistance, while the inhibitory effect of hyperosmolar glucose depended on nutrient-specific feedback on the stomach more than duodenal resistance.