Glyburide (GB) and glipizide (GZ) differ in their pharmacokinetics, but it is not known whether they also differ in mode of action. To examine this question, 10 young healthy subjects and 6 non-insulin-dependent diabetic (NIDDM) patients participated in each of three studies: 1) infusion of saline for 120 min followed by a 100-min hyperglycemic (125 mg/dl) clamp; 2) 120-min primed continuous infusion of GZ followed by a 100-min hyperglycemic clamp; and 3) 120-min primed continuous infusion of GB followed by a 100-min hyperglycemic clamp. The GB and GZ infusions were continued throughout the hyperglycemic clamp. Similar plasma concentrations of GB and GZ were obtained in both groups. All studies were performed with [3-3H]glucose to allow quantification of hepatic glucose production. When administered under basal conditions of glycemia, the acute phase (0-10 min) of plasma insulin and C-peptide increase in both control and NIDDM subjects was twice as great with GZ compared with GB (P less than .01). During the hyperglycemic-clamp studies performed in normal subjects, both GB and GZ increased the first- (1.6-fold) and second- (2.2-fold) phase plasma insulin responses more than hyperglycemia alone. During the hyperglycemic clamp in NIDDM subjects, the first-phase plasma insulin response was absent, and the second-phase insulin response was markedly impaired. Neither GB nor GZ improved first-phase insulin secretion in the NIDDM patients. In both NIDDM and control subjects, the effects of hyperglycemia and sulfonylurea drugs (both GB and GZ) on the first- and second-phase plasma insulin responses were simply additive.(ABSTRACT TRUNCATED AT 250 WORDS)