We tested the hypothesis that a lack of suppression of glucagon causes postprandial hyperglycemia in subjects with type 2 diabetes. Nine diabetic subjects ingested 50 g glucose on two occasions. On both occasions, somatostatin was infused at a rate of 4.3 nmol/kg x min, and insulin was infused in a diabetic insulin profile. On one occasion, glucagon was also infused at a rate of 1.25 ng/kg x min to maintain portal glucagon concentrations constant (nonsuppressed study day). On the other occasion, glucagon infusion was delayed by 2 h to create a transient decrease in glucagon (suppressed study day). Glucagon concentrations on the suppressed study day fell to about 70 ng/L during the first 2 h, rising thereafter to approximately 120 ng/L. In contrast, glucagon concentrations on the nonsuppressed study day remained constant at about 120 ng/L throughout. The decrease in glucagon resulted in substantially lower (P < 0.001) glucose concentrations on the suppressed compared with the nonsuppressed study days (9.2+/-0.7 vs. 10.9+/-0.8 mmol/L) and a lower (P < 0.001) rate of release of [14C]glucose from glycogen (labeled by infusing [1-14C]galactose). On the other hand, flux through the hepatic UDP-glucose pool (and, by implication, glycogen synthesis), measured using the acetaminophen glucuronide method, did not differ on the two occasions. We conclude that lack of suppression of glucagon contributes to postprandial hyperglycemia in subjects with type 2 diabetes at least in part by accelerating glycogenolysis. These data suggest that agents that antagonize glucagon action or secretion are likely to be of value in the treatment of patients with type 2 diabetes.