We previously reported that colestimide, an anion exchange resin, improved glycemic control in patients with type 2 diabetes. However, the factors associated with the decrease of HbA1c remain unclear. In present study, we retrospectively compared glycemic control between groups receiving colestimide (n=71), atorvastatin (n=99), pravastatin (n=85), and pitavastatin (n=95) until 3 months after the start of treatment. In the colestimide group, fasting plasma glucose decreased significantly from 169 ± 59 to 138 ± 29 mg/dL after 3 months (P<0.01), and glycated hemoglobin (HbA1c) declined from 8.1 ± 1.0% to 7.4 ± 0.8% (an 8% reduction, P<0.01). Fasting plasma glucose and HbA1c did not change in the pravastatin and pitavastatin groups. On the other hand, both parameters increased significantly in the atorvastatin group. Multivariate analysis revealed that baseline HbA1c was the main determinant of the decrease of HbA1c in the colestimide group while age, sex, BMI, and baseline lipid levels were not correlated with the effect of colestimide treatment. The decrease of HbA1c showed a positive correlation with baseline HbA1c (r=0.60, P<0.0001), and patients with a larger change of HbA1c (>8.4%) displayed a better response to colestimide. In conclusion, since patients with type 2 diabetes often have hyperlipidemia as well, colestimide therapy may have a clinically useful dual action in such patients. Baseline HbA1c has the most important independent influence on the glucose-lowering effect of colestimide.