The aim of this study was to determine the role of glucagon in hepatic glutamine (Gln) metabolism during exercise. Sampling (artery, portal vein, and hepatic vein) and infusion (vena cava) catheters and flow probes (portal vein, hepatic artery) were implanted in anesthetized dogs. At least 16 days after surgery, an experiment, consisting of a 120-min equilibration period, a 30-min basal sampling period, and a 150-min exercise period, was performed in these animals. [5-¹⁵N]Gln was infused throughout experiments to measure gut and liver Gln kinetics and the incorporation of Gln amide nitrogen into urea. Somatostatin was infused throughout the study. Glucagon was infused at a basal rate until the beginning of exercise, when the rate was either 1) gradually increased to simulate the glucagon response to exercise (n = 5) or 2) unchanged to maintain basal glucagon (n = 5). Insulin was infused during the equilibration and basal periods at rates designed to achieve stable euglycemia. The insulin infusion was reduced in both protocols to simulate the exercise-induced insulin decrement. These studies show that the exercise-induced increase in glucagon is 1) essential for the increase in hepatic Gln uptake and fractional extraction, 2) required for the full increment in ureagenesis, 3) required for the specific transfer of the Gln amide nitrogen to urea, and 4) unrelated to the increase in gut fractional Gln extraction. These data show, by use of the physiological perturbation of exercise, that glucagon is a physiological regulator of hepatic Gln metabolism in vivo.