Objective To compare the efficacy and tolerability of monotherapy with vildagliptin and acarbose in patients with type 2 diabetes (T2DM). Methods Drug-naive T2DM patients were treated by vildagliptin(100 mg daily, given as two equally-divided doses, n=399)or acarbose (up to 300 mg daily, given as three equally-divided doses, n=202) for 24 weeks in this multi-center, randomized, double-blind, double-dummy, parallel-arm and active-controlled study. Results Vildagliptin and acarbose decreased HbA1Cto a similar extent during 24-week treatment. The adjusted mean change from baseline to endpoint in HbA1C was-1.44% in patients receiving vildagliptin and-1.36%in those receiving acarbose, meeting the statistical criterion for non-inferiority (upper limit of 95% confidence interval for between-treatment difference ≤0.4%). There was no significant difference in responder rates between groups(P>0.05). Fasting plasma glucose(FPG) tended to decrease more with acarbose (-1.52 mmol/L)than with vildagliptin(-1. 25 mmol/L), not meeting the statistical criterion for non-inferiority (upper limit of 95% confidence interval for between-treatment difference >0.6 mmol/L). Body weight did not change in vildagliptin-treated patients(-0.5 kg)but decreased in acarbose-treated patients (-1. 8 kg, P<0.01). The proportion of patients experiencing any adverse event was 33. 6% vs 50. 0% in patients receiving vildagliptin or aearbose, respeetively. Adverse gastrointestinal events were significantly more frequent with acarbose(24. 8%)than with vildagliptin (11.3%, P<0.01). No hypoglycemia was reported in either treatment group. Conclusion Vildagliptin is effective and well tolerated in patients with T2DM, demonstrating similar HbA1C reduction as aearbose, but with less adverse gastrointestinal event. Key words: Dipeptidyl peptidase-Ⅳ ; HbA1C ; Glucagon-like peptide-1 ; Incretin hormones