Abstract Aims Obesity is common and associated with unique phenotypic features in heart failure with preserved ejection fraction (HFpEF). Therefore, understanding the efficacy and safety of new therapies in HFpEF patients with obesity is important. The effects of dapagliflozin were examined according to body mass index (BMI) among patients in the Dapagliflozin Evaluation to Improve the LIVEs of Patients With PReserved Ejection Fraction Heart Failure trial. Methods and results Body mass index was analysed by World Health Organization (WHO) categories and as a continuous variable using restricted cubic splines. Body mass index ranged from 15.2 to 50 kg/m2 with a mean value of 29.8 (standard deviation ± 6.1) kg/m2. The proportions, by WHO category, were: normal weight 1343 (21.5%); overweight 2073 (33.1%); Class I obesity 1574 (25.2%); Class II obesity 798 (12.8%); and Class III obesity 415 (6.6%). Compared with placebo, dapagliflozin reduced the risk of the primary outcome to a similar extent across these categories: hazard ratio (95% confidence interval): 0.89 (0.69–1.15), 0.87 (0.70–1.08), 0.74 (0.58–0.93), 0.78 (0.57–1.08), and 0.72 (0.47–1.08), respectively (P-interaction = 0.82). The placebo-corrected change in Kansas City Cardiomyopathy Questionnaire total symptom score with dapagliflozin at 8 months was: 0.9 (−1.1, 2.8), 2.5 (0.8, 4.1), 1.9 (−0.1, 3.8), 2.7 (−0.5, 5.8), and 8.6 (4.0, 13.2) points, respectively (P-interaction = 0.03). The placebo-corrected change in weight at 12 months was: –0.88 (−1.28, –0.47), –0.65 (−1.04, –0.26), –1.42 (−1.89, –0.94), –1.17 (−1.94, –0.40), and –2.50 (−4.4, –0.64) kg (P-interaction = 0.002). Conclusions Obesity is common in patients with HFpEF and is associated with higher rates of heart failure hospitalization and worse health status. Treatment with dapagliflozin improves cardiovascular outcomes across the spectrum of BMI, leads to greater symptom improvement in patients with obesity, compared with those without, and has the additional benefit of causing modest weight loss.