To study whether glucose level in mesenteric artery has some direct effect on the intestinal glucose absorption, the rat small intestine was isolated and vascularly and luminally perfused. The vascular perfusate consisted of Krebs-Ringer bicarbonate buffer (KRB) containing 4% dextran, 0.5% albumin and 1–28 mM glucose, and the luminal perfusate consisted of KRB containing 18 mM glucose. Glucose absorption was estimated from the glucose concentration in the portal venous effluent. The glucose absorption rate was decreased in a dose-dependent manner by the increasing concentration of glucose in the vascular perfusate within a range of 1–28 mM, but not changed by the addition of fructose to the vascular perfusate. It is suggested that glucose in the mesenteric artery has a direct inhibitory effect on the intestinal glucose absorption, not through a change of osmotic press