Fifteen Type 2 diabetics were treated for 4-week periods with once daily (10 mg) glibenclamide, glipizide and placebo according to a double-blind cross-over protocol. Post-dose glipizide concentrations were three times higher than those of glibenclamide, due to the incomplete bioavailability of the latter. On the other hand, pre-dose drug levels were similar, as an expression of the slower absorption and/or elimination of glibenclamide. Both active treatments reduced postprandial blood glucose concentrations and 24-hour urinary glucose excretion to a similar degree, but fasting blood glucose concentrations were slightly lower during glibenclamide treatment. Both active treatments enhanced fasting and postprandial insulin and C-peptide concentrations, the C-peptide response being greater after glipizide than after glibenclamide. Plasma glucagon and GIP concentrations were not significantly affected. Insulin sensitivity was increased by glibenclamide but not by glipizide. Neither therapy affected insulin binding to erythrocytes. It appears that both glibenclamide and glipizide improved glucose metabolism by sustained stimulation of insulin secretion, which was most pronounced with glipizide. Only glibenclamide improved insulin sensitivity and was slightly more active than glipizide on fasting blood glucose levels. The differences may be consequences of the pharmacokinetics, but differences in pharmacodynamics cannot be excluded.