In vitro studies indicate that glucagon-like peptide-1(7-36)-amide (GLP-1) can enhance hepatic glucose uptake. To determine whether GLP-1 increases splanchnic glucose uptake in humans, we studied seven subjects with type 1 diabetes on two occasions. On both occasions, glucose was maintained at approximately 5.5 mmo/l during the night using a variable insulin infusion. On the morning of the study, a somatostatin, glucagon, and growth hormone infusion was started to maintain basal hormone levels. Glucose (containing [3H]glucose) was infused via an intraduodenal tube at a rate of 20 micromol.kg(-1).min(-1). Insulin concentrations were increased to approximately 500 pmol/l while glucose was clamped at approximately 8.8 mmol/l for the next 4 h by means of a variable intravenous glucose infusion labeled with [6,6-2H2]glucose. Surprisingly, the systemic appearance of intraduodenally infused glucose was higher (P = 0.01) during GLP-1 infusion than saline infusion, indicating a lower (P < 0.05) rate of initial splanchnic glucose uptake (1.4 +/- 1.5 vs. 4.8 +/- 0.8 micromol.kg(-1).min(-1)). On the other hand, flux through the hepatic uridine-diphosphate- glucose pool did not differ between study days (14.2 +/- 5.5 vs. 13.0 +/- 4.2 micromol.kg(-1).min(-1)), implying equivalent rates of glycogen synthesis. GLP-1 also impaired (P < 0.05) insulin-induced suppression of endogenous glucose production (6.9 +/- 2.9 vs. 1.3 +/- 1.4 micromol.kg(-1).min(-1)), but caused a time-dependent increase (P < 0.01) in glucose disappearance (93.7 +/- 10.0 vs. 69.3 +/- 6.3 micromol.kg(-1).min(-1); P < 0.01) that was evident only during the final hour of study. We conclude that in the presence of hyperglycemia, hyperinsulinemia, and enterally delivered glucose, GLP-1 increases total body but not splanchnic glucose uptake in humans with type 1 diabetes.