The effects of adding rosiglitazone to existing sulfonylurea (SU) treatment have not previously been studied in Chinese patients with type 2 diabetes and no known pre-existing hepatic impairment. Patients were randomized to receive rosiglitazone 2 mg twice daily (R4 + SU) or 4 mg twice daily (R8 + SU) or placebo (SU + P) for 24 weeks in addition to existing SU treatment. Most patients were taking concomitant glibenclamide (34%) or gliclazide (25%). Changes in glycosylated hemoglobin (HbA(1c)), fasting plasma glucose (FPG), and plasma insulin concentrations were measured. Of the 530 patients enrolled (45% male, mean age 59 years), 105 were in the SU + P group, 215 in the R4 + SU group, and 210 in the R8 + SU group. The mean baseline HbA(1c) was 9.8%, and FPG was 183.8 mg/dL. Compared with placebo, addition of rosiglitazone (2 or 4 mg twice daily) produced significant decreases in mean HbA(1c) (1.04% and 1.44%, respectively; p < 0.0001) and FPG (21.6 and 36.0 mg/dL, respectively; p < 0.0001). There were statistically significant (p < 0.0001) reductions from baseline in insulin concentration of 23.3 and 30.4 pmol/L in the R4 + SU and R8 + SU groups, respectively. Despite the high prevalence of seropositivity for hepatitis B and/or C at baseline (56%), there was no evidence of hepatotoxicity. No clinically significant changes in routine hematology, biochemistry, or electrocardiogram were observed. The addition of rosiglitazone to SU produced clinically significant improvements in glycemic control in Chinese patients with type 2 diabetes. Rosiglitazone plus SU was well tolerated irrespective of hepatitis B and C serological status.