Abstract Background SGLT2 inhibitors have been shown to reduce hospitalization for heart failure (HHF). We sought to determine independent baseline predictors for HHF specifically in a population with type 2 diabetes and chronic kidney disease (CKD). Methods CREDENCE randomized 4401 participants with type 2 diabetes and CKD to canagliflozin 100 mg versus placebo. We evaluated the baseline clinical and demographic factors using multivariate regression modeling to identify the independent predictors of HHF. Results Overall, 230 participants (89 canagliflozin; 141 placebo) had at least 1 HHF event. Canagliflozin reduced the incidence of HHF compared with placebo (4.0% vs 6.4%; HR 0.61; 95% CI 0.47–0.80). Participants with HHF events postrandomization were older (65.8 vs 62.9 y), and had a longer duration of diabetes (17.4 vs 15.7 y), higher prevalence of prior HF (30.4% vs 14.0%), higher urinary albumin:creatinine ratio (1347 vs 904 mg/g), lower estimated glomerular filtration rate (51.5 vs 56.4 mL/min/1.73m2), and higher prevalence of prior cardiovascular disease (65.7% vs 49.6%) compared to those without HHF. Independent predictors of HHF are shown in the Table. Conclusions HHF is common in patients with type 2 diabetes and CKD. Canagliflozin reduces HHF by 39% compared with placebo. Higher urinary albumin:creatinine ratio was the most potent predictor of HHF and should be part of patient risk assessment. Funding Acknowledgement Type of funding source: Private company. Main funding source(s): Janssen Research & Development, LLC