Arterial dysfunction (AD) in type 2 diabetes mellitus (T2DM) predicts cardiovascular events. The objective was to investigate the prevalence and predictors of AD in statin-treated T2DM patients. We measured flow-mediated (FMD) and nitrate-mediated (NMD) brachial artery dilatation in 86 statin-treated T2DM patients. Patients were classified into 2 groups: normal arterial function (FMD ≥3.7% with NMD ≥11.9%) or AD (FMD <3.7% with or without NMD <11.9%). Endothelial dysfunction without smooth muscle cell dysfunction (ED) was defined as FMD less than 3.7% with NMD of at least 11.9%, and endothelial dysfunction with smooth muscle cell dysfunction (ED/SMD) was defined as FMD less than 3.7% with NMD less than 11.9%. Predictors of arterial function were investigated using linear and logistic regression methods. The prevalence of AD was 33.7% (23.2% with ED and 10.5% with ED/SMD). In multivariate linear regression, history of hypertension (P < .01), statin dose (P < .05), and estimated glomerular filtration rate (eGFR) (P = .02) were significant predictors of FMD. Sex (P < .01) and creatinine (P = .03) or eGFR (P = .02) predicted NMD. In multivariate logistic regression, the independent predictors of AD were history of hypertension (odds ratio [OR], 8.79; 95% confidence interval, 2.14-36.12; P < .01), age (OR, 1.08; 1.01-1.17; P = .03), and statin dose (OR, 0.33; 0.12-0.87; P = .02). A history of hypertension (OR, 8.99; 1.87-43.26; P < .01) was the sole independent predictor of ED; eGFR (OR, 0.01; 0.00-0.26; P < .01) independently predicted ED/SMD. Our data suggest that one third of statin-treated diabetic patients have residual AD, mainly due to ED alone. Earlier identification and treatment of hypertension and renal impairment may improve AD and further decrease cardiovascular risk in such patients.