OBJECTIVES: To evaluate relative efficacy and safety of SGLT2 inhibitor compared to DPP4 inhibitor in patients with T2DM in the absence of head-to-head evidence. METHODS: A systematic search of the literature was conducted in MEDLINE, EMBASE, LILACS, the Cochrane Central Register of Controlled Trials, and Clinicaltrials.gov to June 2015. Selected studies were randomized controlled trials (RCTs) that compare DPP4i plus insulin (DPP4i/INS) or SGLT2i plus insulin (SGLT2i/INS) with placebo plus insulin (PCB/INS), as a common comparator, in patients with T2DM. The primary outcome was the change in HbA1c from baseline to the end of the intervention period. The secondary outcomes included the change in fasting plasma glucose, the change in body weight, and the event of hypoglycemia. Covariate-adjusted indirect comparison using  Bayesian meta-regression analyses was performed on WinBUGS14. RESULTS: We included 14 eligible RCTS comprising 6,980 patients, of which 9 were DPP4i studies and 5 were SGLT2i studies. Unadjusted indirect comparison analyses suggested that the HbA1c reduction between SGLT2i/INS and DPP4i/INS groups did not differ (WMD -0.07%, 95% CI* -0.26 to 0.12%). However, covariate-adjusted analyses showed that SGLT2i/INS achieved greater reduction in HbA1c (WMD -0.24%, 95% CI -0.44 to -0.03%). Covariates included in the model were age, sex, BMI and baseline insulin dose. For the secondary outcomes, the results showed that SGLT2i/INS achieved greater reduction in fasting plasma glucose (WMD -18.8mg/dL, 95% CI -28.1 to -9.4 mg/dL) and body weight (WMD -2.39 kg, 95% CI -3.12 to -1.66 kg) than DPP4i/INS without increasing the risk of hypoglycemia (RR 1.19, 95% CI 0.80 to 1.78). *CI : Credible Interval CONCLUSIONS: The addition of SGLT2i on top of pre-existing insulin therapy suggests a potential for a better glycemic control and greater weight reduction than DPP4i in patients with inadequately controlled T2DM.