In this study, we characterized type 2 diabetic patients responding well to the alpha-glucosidase inhibitor voglibose administration as an adjunct to sulfonylurea treatment. Thirty-three type 2 diabetic patients were enrolled in an open prospective study. All the patients had been treated for at least 1 year with a sulfonylurea drug, in whom HbA1c level had been stable for at least 12 weeks. The patients were given voglibose at a dose of 0.2 mg t.i.d. for 12 weeks. Voglibose administration significantly decreased the mean HbA1c level in all the patients at 4, 8, and 12 weeks. Twelve (36%) of the study patients were responders, when the responders were defined as patients whose HbA1c level at 12 weeks fell by at least 1.0% from baseline, or those whose HbA1c level at 12 weeks was < or =7.0%, falling by at least 0.5% from baseline. The baseline fasting plasma glucose (FPG) was significantly lower, and the baseline homeostasis model assessment (HOMA) beta-cell function (HOMA-%beta) was significantly higher in the responders than in the non-responders. There were more patients who had FPG <170 mg/dl and/or HOMA-%beta > or =30% in the responders than in the non-responders (P<0.005). None of the patients with both FPG > or =170 mg/dl and HOMA-%beta >30% responded to the adjunct treatment. These results indicate that baseline FPG and HOMA-%beta are useful clinical markers to predict the effectiveness of the adjunct therapy of voglibose in sulfonylurea-treated type 2 diabetic patients.