To evaluate the role of glucagon in insulin-mediated glucose metabolism, we studied four men and four women, ranging in age from 30-73 yr (mean +/- SEM, 54 +/- 5) who had undergone complete pancreatic resection for cancer or chronic pancreatitis 16-58 mo previously. The patients had undetectable C-peptide levels and established lack of biologically active 3500 mol wt glucagon. Euglycemic insulin clamp studies were performed with a 40 mU X m-2 X min-1 insulin infusion in the basal, post-absorptive, insulin-withdrawn state, before and during the last 3 h of a 72-h glucagon replacement-dose infusion (1.25 ng X kg-1 X min-1). In four patients, hepatic glucose production was determined by a primed-constant infusion of 3-[3H]glucose. Monocyte insulin-binding studies, pre- and postglucagon, were performed in all patients. The 72-h glucagon infusion, resulting in mean plasma glucagon levels of 124 +/- 7 pg/ml, caused a significant rise in the mean plasma glucose level (249 +/- 8 versus 170 +/- 13 mg/dl preglucagon) and a sixfold increase in mean 24-h glucose excretion. Both with and without glucagon, euglycemic hyperinsulinemia achieved identical and complete suppression of hepatic glucose production. The mean glucose utilization rate (4.70 +/- 0.36 mg X kg-1 X min-1 preglucagon) was significantly decreased by glucagon replacement (3.83 +/- 0.31 mg X kg-1 X min-1, P less than 0.02). Mean glucose clearance was also diminished with glucagon (4.49 +/- 0.32 versus 5.73 +/- 0.45 ml X kg-1 X min-1 preglucagon, P less than 0.02).(ABSTRACT TRUNCATED AT 250 WORDS)