〔Abstract〕 Objective To investigate the correlation between diabetic kidney disease and sweat function by measuring urine albumin to creatinine ratio(UACR) and the skin electrical conductivity(ESC). Methods A total of 300 type 2 diabetes mellitus(T2 DM) patients were enrolled in this study. UACR and Sudoscan were done for each patient. The conductivity analyzer Sudoscan was used to measure the skin electrical conductivity(ESC) of hands and feet. All patients were divided into two groups according to the hands ESC(HESC) and feet ESC(FESC), i.e. patients with normal sweat function(HESC>60 μs and FESC>70 μs, NSF group,n=125) and patients with abnormal sweat function(HESC≤60 μs or FESC≤70 μs, ASF group,n=175). The relevance of UACR to sweat function was later analyzed. Results Age, the duration of diabetes and UACR were significantly higher in the ASF group compared with the NSF group, while HESC and FESC were significantly lower(P<0. 05 orP<0. 01). Spearman correlation analysis showed that both HESC and FESC negatively correlated with UACR(rs=-0.16,-0.17,P<0.01), the duration of diabetes(rs=-0.18,-0.15,P<0.01) and Scr(rs=-0.11,-0.16,P<0.05). Multivariate stepwise linear regression analysis showed that the duration of diabetes and UACR were independent risk factors for HESC. The duration of diabetes, UACR and Scr were independent risk factors for FESC. When UACR was used as the criterion for diagnosing diabetic kidney disease, the area under curve(AUC) of HESC and FESC was 58.7% and 59.8% respectively, the specificity was 64.3% and 70.5% respectively, the sensitivity was 57.2% and 65.6% respectively. Conclusion Sweat function test is a potential predictor for kidney disease in patients with T2 DM.