Increased urinary albumin excretion in diabetes not only signals nephropathy but also serves as a risk marker for cardiovascular disease. The data of MARCH (Metformin and AcaRbose in Chinese as the initial Hypoglycaemic treatment) trial demonstrated that acarbose and metformin were similarly efficacious at lowering blood glucose and blood pressure, as well as improving insulin sensitivity in Chinese patients newly diagnosed with type 2 diabetes mellitus. The purpose of this study was to identify the effects of acarbose and metformin therapy on albumin excretion in MARCH study.Baseline urine albumin/creatinine ratio (ACR) of 762 newly diagnosed, drug-naive patients with type 2 diabetes mellitus was measured. Included patients were randomized to receive either acarbose or metformin and followed for 48 weeks. In addition to change in ACR, the estimated glomerular filtration rates (eGFR) and frequency of metabolic syndrome (MetS) were also assessed.Elevated ACR levels (30mg/g) were present at baseline in 21.9% of all participants. A significant decline in urine ACR was observed in both the acarbose and metformin groups at week 24 and 48 (all P<0.001). The proportion of patients with elevated ACRs was also reduced in both treatment groups at week 24 and 48 compared with baseline values (all P<0.05). The change in urine ACR at week 48 was significantly greater in patients prescribed acarbose than in those prescribed metformin (P=0.01). Both acarbose and metformin significantly decreased the frequency of MetS at week 24 and 48 (both P<0.05). Neither treatment affected eGFR.In sum, both acarbose and metformin decreased urine ACR levels and reduced the frequency of elevated ACR and MetS in Chinese patients with newly diagnosed type 2 diabetes mellitus without affecting eGFR. After 48 weeks' intervention, acarbose therapy resulted in a greater reduction in urine ACR compared with metformin.