Purpose To investigate the reliability of 1,5-anhydroglucitol (1,5-AG) in diabetes with mild or moderate renal dysfunction. Methods 668 patients diagnosed with diabetes as DM group and 336 healthy controls as non-DM group were enrolled in this study. DM group was divided into fou r groups according to estimated glo-merularfiltration rate (eGFR). Serum concentrations of 1,5-AG, fructosamine (FMN) and glycated hemoglobin (HbA1c) were assayed via the enzymatic method, the nitro reduction four nitrogen thiazole blue test,high performance liquid chromatography respectively. Results In diabetic Patients with eGFR >= 30 mL/min, significant negative association still existed between logarithmic transformed 1,5-AG values (In1,5-AG) and HbAl c (all P<0.001) as well as fasting plasma glucose (FPG) (all P<0.05). Besides, eGFR was not one of the determinants of 1,5-AG levels in both DM group (standard beta= - 0.049, P=0.383 ) and healthy controls (r = -0.095, P = 0.084 ). Stepwise multiple linear regression showed that serum uric acid (UA) is one of the influencing factors of 1,5-AG (standard beta = 0.119, P=0.015) and among the three glycemic markers, only HbA1c was found to be correlated with the homeostasis model assessment for b-cell function (HOMA-beta) (standard beta= 0.097, P=0.012). Conclusions 1,5-AG values remain reliable as a glycemic control marker In diabetes with mild or moderate dysfunction. Serum UA was significantly and positively correlated with 1,5AG levels. HbA1c may be a good biomarker for insulin resistance compared with 1,5-AG and FMN.