Background: A Bayesian network meta-analysis (BNMA) was conducted to systematically evaluate efficacy and safety of sodium-glucose co-transporter-2 inhibitors (SGLT2is) in treating DKD. Methods: PubMed, EMBASE, Web of Science, Cochrane Library and ClinicalTrials.gov were searched from inception to Mar 23, 2025. Randomized clinical trials (RCTs) of SGLT2is in treating DKD were included. 17 efficacy and 13 safety outcomes were included, e.g., HbA1c, eGFR, UACR, UTI, GMI. Results: 17 studies with 22,774 participants were included. Among all the included interventions, 10 mg of Dap + 2.5 mg of Sax demonstrated superior efficacy in reducing HbA1c, followed by 25 mg of Emp, 400 mg of Sot, 10 mg of Emp, 20 mg of Bex; 100 mg of Can showed the best efficacy in increasing eGFR, followed by placebo, 20 mg of Bex, 200 mg of Sot, 400 mg of Sot; 300 mg of Can showed the best efficacy in reducing UACR, followed by 25 mg of Emp, 100 mg of Can 10 mg of Dap + 2.5 mg of Sax, 10 mg of Dap; 100 mg of Can showed the lowest incidence of UTI, followed by 5 mg of Dap, 400 mg of Sot, 10 mg of Dap, 5 mg of Ert. Conclusion: SGLT2is displayed favorable efficacy and acceptable safety in treating DKD. 10 mg of Dap + 2.5 mg of Sax demonstrated superior efficacy in reducing HbA1c. 100 mg of Can exhibited the optimal efficacy in improving eGFR and safety. 300 mg of Can showed the best efficacy in reducing UACR.