Hyperinsulinemia and peripheral insulin resistance caused by systemic insulin delivery and prednisone therapy are recognized consequences of pancreas transplantation. However, there is little information about insulin action on the liver. To investigate hepatic insulin sensitivity in recipients of pancreas transplants, we devised a staged euglycemic hyperinsulinemic clamp to measure hepatic glucose production (HGP) in 10 type I diabetic pancreas transplant recipients, 10 pair-matched healthy control subjects, and 6 nondiabetic kidney transplant recipients. Clamps were performed in two sequential stages. In stage 1, a 2-h low-dose insulin infusion (0.4 mU · kg-1 · min-1) was used to partially suppress HGP. In stage 2, insulin- mediated suppression of HGP was challenged by a 1.5-h glucagon infusion (0.8 ng · kg-1 · min-1), while continuing the hyperinsulinemic euglycemic- clamp conditions. During both stages, somatostatin (250 μg/h) was infused to suppress endogenous insulin secretion. All subjects underwent stage 1, and all except one pancreas recipient and a respective matched healthy control subject completed stage 2. Fasting HGP was greater in pancreas recipients than in healthy control subjects (15.1 ± 0.7 vs. 12.0 ± 0.4 μmol · l-1 · kg-1 · min-1, P < 0.005) but similar in healthy control subjects and in kidney recipients. During stage 2, both total (706 ± 28 vs. 469 ± 31 μmol · l-1 · kg-1, P < 0.005) and incremental (62 ± 20 vs. -21 ± 16 μmol · l-1 · kg-1, P < 0.005) HGP responses to glucagon infusion were significantly greater in pancreas recipients than in healthy control subjects. Changes in HGP in kidney recipients during stage 2 were not significantly different from those in healthy control subjects. In conclusion, fasting HGP is increased in pancreas transplant recipients. Furthermore, recipients have hepatic insulin resistance as demonstrated by an enhanced stimulatory effect of glucagon on HGP during insulin-mediated HGP suppression. Because the magnitude of hepatic insulin resistance was a significant (P < 0.01) predictor of HbA(1c) level, we suggest that variable hepatic insulin resistance may be responsible for some of the variance observed in glycemic levels after successful pancreas transplantation.