Abstract Background and Aims In the FAGOTTO study (jRCTs041200069) to evaluate changes of renal hemodynamics on initial glomerular filtration rate (GFR) dip induced by Sodium-glucose cotransporter 2 inhibitors (SGLT2i) in patients (pts) with type 2 diabetic kidney disease (DKD) and moderate renal dysfunction, we reported that SGLT2 inhibition maintained the levels of glomerular filtration fraction (FF) and renal vascular resistance (RVR) during decrease in GFR as well as renal plasma and blood flow. In this sub analysis, we evaluated if GFR dip by SGLT2i might induce tubulointerstitial disorder by measuring urinary biomarkers. Method The FAGOTTO study was a 12-week multicenter, open-label, randomized (1:1), parallel-group study treated with Canagliflozin (Group CA) or control groups (Group CO) in DKD pts with 30 ≤ estimated GFR (eGFR) ≤ 60 mL/min/1.73 m2. Urinary biomarkers including liver-type fatty-acid binding protein (L-FABP), N-acetyl-glucosaminidase (NAG), β2-microglobulin (β2MG) and monocyte chemoattractant protein-1 (MCP-1) were measured at baseline, 4 weeks and 12 weeks after start of the trial. The adjusted mean and SE for changes were calculated at each time point by a linear mixed model with baseline values, time point, and interaction of the treatment group as the fixed effects. We compared these changes between treatment the groups the Tukey-Kramer method to correct for multiplicity. Results The adjusted mean changes in L-FABP at the 4-week and 12-week time point after trial start were estimated −0.35 and −0.17 μg/gCr in Group CA and 0.06 and 1.23 μg/gCr in Group CO in a linear mixed model, respectively. Similarly, the changes were estimated 2.63 and 3.71 U/gCr in Group CA and 1.38 and −1.21 U/gCr in Group CO in NAG, 317.7 and 291.4 μg/gCr in Group CA and 416.6 and 375.1 μg/gCr in Group CO in β2MG and 19.6 and 64.6 pg/gCr in Group CA and −7.7 and −2.7 pg/gCr in Group CO in MCP-1, respectively. When compared the changes in L-FABP, NAG, β2MG and MCP-1 between Group Ca and Group Co by Tukey-Kramer multiple-comparison test, there were no significant differences in any biomarkers. (Figure). Conclusion In this analysis, no significant difference in any of these biomarkers was detected in Group CA compared to those in Group CO at 4 weeks and 12 weeks after administration of SGLT2i. Although further studies should be needed to conclude, transient decrease in GFR after administration of SGLT2i might not be involved with tubulointerstitial damage and our study gives one evidence as a safety information in use of SGLT2i to DKD pts with moderate renal dysfunction.