OBJECTIVE To assess the hypoglycemic effect and the insulin-releasing effect of the main glyburide (glibenclamide) metabolites 4-trans-hydroxy-glibenclamide (M1) and 3-cis-hydroxy-glibenclamide (M2) in humans. METHODS Eight healthy subjects participated in a placebo-controlled, randomized, single-blind crossover study with five single-dose tests, 3 months apart: 3.5 mg glibenclamide (Gb) orally, 3.5 mg Gb intravenously, 3.5 mg M1 intravenously, 3.5 mg M2 intravenously, and placebo intravenously, each in the fasting state. Standardized meals were given 0.5 and 5.5 h after each medication. Blood glucose levels were measured by a glucose oxidase method, and serum insulin concentrations were analyzed by a specific immunoassay. RESULTS Blood glucose levels during the first 5 h were significantly lowered not only by Gb but also by M1 and M2. The mean +/- SE blood glucose reductions (versus placebo) expressed as percent of area under the curve (AUC) (0-5 h) were 18.2 +/- 3.3% for M1, 12.5 +/- 2.3% for M2, 19.9 +/- 2.1% for intravenous Gb, and 23.8 +/- 1.2% for Gb orally. Serum insulin levels were significantly increased by Gb as well as by M1 and M2. and M2. CONCLUSIONS The two main metabolites of glyburide (glibenclamide) have a hypoglycemic effect in humans, which is due to increased insulin secretion.