Six awake adult dogs prepared with a duodenocutaneous fistula were infused intraduodenally with one of the following solutions: 3% saline, 10% glucose, 20% glucose, 20% galactose, 20% fructose, 20% mannose, 20% sorbitol, 20% maltose, 20% lactose, or 20% sucrose. Both 10 and 20% glucose stimulated GIP release, and the response appeared to be dose related. Actively transported galactose (C-4 epimer) stimulated GIP release, but less than glucose. Fructose (C-2 keto sugar) which is absorbed by facilitated transport did not stimulate GIP release. Mannose (C-2 epimer) which is passively absorbed by diffusion did not release GIP. Sorbitol (reduced alcohol of glucose) which is not absorbed did not release GIP. Of the disaccharides tested, only maltose stimulated the release of GIP. The results suggest that structural integrity of the glucose molecule from the C-1 to C-4 carbon atoms, a free aldehyde group on the C-1 carbon atom, and a cyclic structure are all necessary for both the active transport of glucose and the release of endogenous GIP. It would appear that structurally similar receptors exist for both the active transport of glucose and for the release of GIP.