Heart failure with preserved ejection fraction (HFpEF) is often characterized by nitric oxide deficiency.To evaluate the efficacy and adverse effects of praliciguat, an oral soluble guanylate cyclase stimulator, in patients with HFpEF.CAPACITY HFpEF was a randomized, double-blind, placebo-controlled, phase 2 trial. Fifty-nine sites enrolled 196 patients with heart failure and an ejection fraction of at least 40%, impaired peak rate of oxygen consumption (peak V̇o2), and at least 2 conditions associated with nitric oxide deficiency (diabetes, hypertension, obesity, or advanced age). The trial randomized patients to 1 of 3 praliciguat dose groups or a placebo group, but was refocused early to a comparison of the 40-mg praliciguat dose vs placebo. Participants were enrolled from November 15, 2017, to April 30, 2019, with final follow-up on August 19, 2019.Patients were randomized to receive 12 weeks of treatment with 40 mg of praliciguat daily (n = 91) or placebo (n = 90).The primary efficacy end point was the change from baseline in peak V̇o2 in patients who completed at least 8 weeks of assigned dosing. Secondary end points included the change from baseline in 6-minute walk test distance and in ventilatory efficiency (ventilation/carbon dioxide production slope). The primary adverse event end point was the incidence of treatment-emergent adverse events (TEAEs).Among 181 patients (mean [SD] age, 70  years; 75 [41%] women), 155 (86%) completed the trial. In the placebo (n = 78) and praliciguat (n = 65) groups, changes in peak V̇o2 were 0.04 mL/kg/min (95% CI, -0.49 to 0.56) and -0.26 mL/kg/min (95% CI, -0.83 to 0.31), respectively; the placebo-adjusted least-squares between-group difference in mean change from baseline was -0.30 mL/kg/min ([95% CI, -0.95 to 0.35]; P = .37). None of the 3 prespecified secondary end points were statistically significant. In the placebo and praliciguat groups, changes in 6-minute walk test distance were 58.1 m (95% CI, 26.1-90.1) and 41.4 m (95% CI, 8.2-74.5), respectively; the placebo-adjusted least-squares between-group difference in mean change from baseline was -16.7 m (95% CI, -47.4 to 13.9). In the placebo and praliciguat groups, the placebo-adjusted least-squares between-group difference in mean change in ventilation/carbon dioxide production slope was -0.3 (95% CI, -1.6 to 1.0). There were more dizziness (9.9% vs 1.1%), hypotension (8.8% vs 0%), and headache (11% vs 6.7%) TEAEs with praliciguat compared with placebo. The frequency of serious TEAEs was similar between the groups (10% in the praliciguat group and 11% in the placebo group).Among patients with HFpEF, the soluble guanylate cyclase stimulator praliciguat, compared with placebo, did not significantly improve peak V̇o2 from baseline to week 12. These findings do not support the use of praliciguat in patients with HFpEF.ClinicalTrials.gov Identifier: NCT03254485.