NIDDM is associated with excessive rates of endogenous glucose production in both the postabsorptive and postprandial states. To determine whether this is due to an intrinsic increase in hepatic sensitivity to glucagon, 9 NIDDM and 10 nondiabetic subjects were studied on three occasions. On each occasion, glycogen was labeled the evening before the study with subjects ingesting meals containing [6-3H]galactose. Beginning at 6:00 A.M. on the following morning, somatostatin was infused to inhibit endogenous hormone secretion. Insulin concentrations were maintained constant at basal levels (defined as that necessary to keep glucose at approximately 5 mmol/l) in each individual. On one occasion, glucagon was infused at a rate of 0.65 ng x kg(-1) x min(-1) throughout the experiment, resulting in glucagon concentrations of approximately 130 pg/ml and a slow but comparable fall in endogenous glucose production with time in both groups. On the other two occasions, the glucagon infusion was increased at 10:00 A.M. to either 1.5 or 3.0 ng x kg(-1) x min(-1), resulting in an increase in glucagon concentrations to approximately 180 and 310 pg/ml, respectively. The increment in endogenous glucose production (i.e., area above basal) did not differ in diabetic and nondiabetic subjects during either the 1.5 ng x kg(-1) x min(-1) (0.75 +/- 0.055 vs. 0.78 +/- 0.048 mmol/kg) or 3.0 ng x kg(-1) x min(-1) (1.06 +/- 0.066 vs. 1.10 +/- 0.073 mmol/kg) glucagon infusions. In contrast, the amount of [6-3H]glucose released from glycogen was lower (P < 0.05) in the diabetic than nondiabetic subjects during both glucagon infusions. The specific activity of glycogen, calculated as the integrated release of [6-3H]glucose divided by the integrated release of unlabeled glucose, was lower (P < 0.05) in diabetic subjects than in nondiabetic subjects during both the 1.5 ng x kg(-1) x min(-1) (19.0 +/- 3.9 vs. 41.4 +/- 5.7 dpm/micromol) and 3.0 ng x kg(-1) x min(-1) (19.1 +/- 3.1 vs. 36.5 +/- 7.2 dpm/micromol) glucagon infusions, implying that a greater portion of the glucose released from glycogen was derived from the indirect pathway. We concluded that although NIDDM is not associated with an intrinsic alteration in hepatic sensitivity to glucagon, it does alter the relative contributions of the direct and indirect pathways to nocturnal glycogen synthesis.