OBJECTIVES The aim of the present network meta-analysis is to compare the efficacy and safety of empagliflozin versus other antidiabetic drugs used in third line for the treatment of patients with type 2 diabetes mellitus (T2DM). METHODS We conducted a systematic review randomized controlled trials (RCTs) and Bayesian network meta-analysis to establish the comparative efficacy and safety of SGLT-2s, DPP-4s, GLP-1s, and TZDs. RCTs enrolling subjects with T2DM inadequately controlled on metformin plus sulfonylurea were included. The principal outcome of this analysis was the effect of these drugs on HbA1c, weight, systolic blood pressure (SBP), incidence of hypoglycaemia and urinary tract infections (UTIs) at 24 weeks.  RESULTS From 6969 abstracts, 13 were included in the analysis. No RCTs involving TZDs were identified. Compared with placebo, mean changes in HbA1c were -0.65% [95% confidence interval (CI) -1.59 to -0.08%] and -0.60% [95%CI -1.14 to -0.14%] for empagliflozin 10mg and 25mg. No significant differences were detected between interventions. Mean changes in weight with empagliflozin 10mg and 25mg were -1.77 [95%CI -2.19 to -1.35) and -2.00 (95%CI -2.44 to -1.57), respectively. Mean weight losses were fairly similar across SLGT-2s and GLP-1s ranging between -1.26 to -2.12. All DPP-4s were associated weight gains, ranging form 0.33 to 0.98, of which most were statistically significant. SBP data were only available for SGLT-2s and DPP-4s. Empagliflozin 10mg and 25mg compared with placebo had statistically significant reductions of -2.70 and -2.09. All interventions (except exenatide) yielded relative risk of hypoglycaemia greater than 1.00. For UTIs, no differences were found between SLGT-2s or DPP-4s and placebo. CONCLUSIONS Compared with other SGLT-2s, DPP-4s, and GLP-1s, empagliflozin generally offers similar HbA1c control at week 24, an advantageous profile in weight loss and reduction of SBP, as well as similar safety profile.