Alpha-glucosidase inhibitors (AGIs) was reported to be associated with several rare adverse hepatic events, but with inconsistent results. We aimed to investigate the risk of hepatotoxicity associated with the use of AGIs in patients with type 2 diabetes mellitus (T2DM), and performed a systematic review and meta-analysis. Fourteen studies (n = 2881) were eligible, all of which were RCTs. Meta-analysis of data regarding elevation of more than 3-fold the upper limit of normal (ULN) of AST and ALT showed statistically significant differences between AGIs treatment versus control (OR 6.86, 95% CI 2.50 to 18.80; OR 6.48, 95% CI 2.40 to 17.49). Subgroup analyses of elevation of more than 1.8-fold ULN of AST and ALT by dose of AGIs showed differential effects on AST and ALT (AST: OR 0.38 vs 7.31, interaction P = 0.003; ALT: OR 0.32 vs 4.55, interaction p = 0.02). Meta-analysis showed that AGIs might increase the risk of hepatotoxicity, and higher dose appeared to be associated with higher risk of hepatotoxicity. However, the evidence is limited with surrogate measures (i.e. ALT and AST), and no clinically important adverse events were observed.Alpha-glucosidase inhibitors (AGIs) are commonly used oral hypoglycemic drugs, especially in the patient population from East Asia 1-3 . The guideline of the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD) recommended the use of AGIs as a potentially first-line agent or in combination with other antihyperglycemic drugs 4 . AGIs has proven similarly efficacious as other commonly used antidiabetes agents [5][6][7] . A recent large trial 1 showed that acarbose is similar to metformin in terms of efficacy, and supports a viable choice for initial therapy in patients with newly diagnosed type 2 diabetes. Additionally, AGIs do not increase body weight, rarely cause hypoglycemia; and have minimal drug-drug interactions 1,7,8 . Meanwhile, AGIs was reported to be associated with several rare adverse hepatic events 9-11 and increase liver enzyme levels [12][13][14][15][16][17][18] . The causal relationship, however, has not been established 9 , and the magnitude of effect on the increase of liver enzyme levels remains unclear. Because these issues are often treated as adverse effects issues, and the hepatic adverse events, if any, are usually rare, individual trials are not adequate to address these important clinical questions. A meta-analysis -in which multiple studies are pooled -may offer opportunity to detect a small but clinically important difference.Thus, we carried out a systematic review of randomized controlled trials and observational studies to assess the association between hepatotoxicity and AGIs. We hypothesized that hepatotoxicity would be more frequently manifested in AGIs as opposed to no use. Figure 1 showed the study selection process. We acquired 5,318 reports. After title and abstract screening, 178 were potentially eligible (including 159 potentially relevant RCTs and 19 potentially relevant observational studies [1...