BackgroundSodium‐glucose cotransporter‐2 inhibitors (SGLT2i) are recommended by the American Heart Association for management of heart failure with preserved ejection fraction (HFpEF), but little is known about their in‐class comparative effectiveness in real‐world settings.ObjectivesTo assess the in‐class comparative effectiveness of SGLT2i for preventing HF‐related and all‐cause hospitalizations among patients with HFpEF.MethodsUsing MarketScan® Commercial and Medicare Supplemental research databases (2012–2020), this cohort study included adults with HFpEF treated with SGLT2i. Stabilized inverse probability treatment weighted Cox proportional hazards regression was used to compare HF‐related and all‐cause hospitalizations in three pairwise comparisons: dapagliflozin versus canagliflozin, empagliflozin versus canagliflozin, and dapagliflozin versus empagliflozin. Subgroup and sensitivity analyses were conducted to assess robustness of the main analysis.ResultsIn total, 3629 SGLT2i users (881 dapagliflozin, 1120 canagliflozin, and 1628 empagliflozin) were included. Compared with canagliflozin, dapagliflozin was associated with decreased risk of HF‐related hospitalization (adjusted hazard ratio [aHR], 0.75; 95% confidence interval [CI], 0.56–1.01) and all‐cause hospitalization (aHR, 0.84; 95% CI 0.73–0.97). Compared with canagliflozin, empagliflozin was associated with 55% decreased risk of HF‐related hospitalization (aHR, 0.45; 95% CI 0.34–0.59) and 18% decreased risk of all‐cause hospitalization (aHR, 0.82; 95% CI 0.73–0.93). Compared with empagliflozin, dapagliflozin was associated with 50% increased risk of HF‐related hospitalization (aHR, 1.50; 95% CI 1.09–2.07) and a statistically nonsignificant increase in the risk of all‐cause hospitalization (aHR, 1.05; 95% CI 0.92–1.20).ConclusionsCompared with canagliflozin or dapagliflozin use, empagliflozin use was associated with decreased risk of HF‐related and all‐cause hospitalizations. Compared with canagliflozin, dapagliflozin was associated with a reduced risk of HF‐related and all‐cause hospitalizations.