The aim of the present study was to describe the mechanism by which the combination glyburide/metformin exerts its additive hypoglycemic effects. This is a double-blind, randomized and crossover clinical trial. Patients (n = 20) were included in a run-in period of 8 weeks in which an isocaloric diet was prescribed. If they did not achieve the treatment goals (n = 15), they received glyburide, metformin or combined treatment for 10 weeks each using three possible sequences. The dosage was adjusted to reach fasting plasma glucose (FPG) < 7.7 mmol/l. Treatment periods were separated by a 6-12 week washout period. At the beginning and the end of every treatment, insulin sensitivity and insulin secretion were measured by means of a minimal model and an oral glucose tolerance test. All treatment periods were completed by 12 cases. The glycemic goal was reached in 1 case during metformin, in 5 during glyburide and in 10 during the combination. The greatest reduction in HbA1c was achieved during the combination (HbA1c 11 +/- 1.6 vs 9.8 +/- 1.9 vs 9.0 +/- 2.1% for metformin, glyburide and the combination, p < 0.001). Increased insulin secretion was the explanation for the additive effects of the combination (percentual change in acute insulin response during the minimal model = 5.8 vs 51.5 vs 88.2% for metformin, glyburide and the combination, p < 0.05). No change in insulin sensitivity resulted from the treatments. In conclusion, the additive hypoglycemic effects of the combination glyburide/metformin was caused by increased insulin secretion.