Objectives The objective of the study is to compare the efficacy of hypoglycaemic drugs for type 2 diabetes mellitus (T2DM) by network meta-analysis of randomized controlled trials (RCTs). Methods We compared 11 major oral hypoglycaemic drugs under five categories evaluated by RCTs as drug monotherapy for the patients with T2DM, measuring glycosylated haemoglobin (%) or fasting plasma glucose (mmol L-1) as outcomes. RCT quality was assessed with the Cochrane risk of bias tool. Network meta-analysis estimated the mean differences and 95% credible intervals. Subgroup and sensitivity analyses were performed to determine the results robustness. The Grading of Recommendation, Assessment, Development, and Evaluation evidence strength was assessed. Results Seventy-five RCTs including 33,830 patients were identified. Their study quality was high. Regarding glycosylated haemoglobin, top three anti-diabetics were repaglinide (mean differences -1.39 [95% credible intervals -1.75 to -1.03]), gliclazide (-1.37 [-2.04 to -0.71]) and metformin (-1.13 [-1.37 to -0.90]), against placebo. Regarding fasting plasma glucose, top three anti-diabetics were repaglinide (-2.01 [-2.75 to -0.97]), metformin (-1.72 [-2.16 to -1.27]) and glipizide (-1.57 [-2.44 to -0.64]), against placebo. There was no difference between metformin and repaglinide. Subgroup and sensitivity analyses found the results to be robust. The evidence strength was moderate to high. Conclusion This meta-analysis showed that repaglinide and metformin would be the most efficacious oral drugs for first-line monotherapy of T2DM.