Half of patients with heart failure with preserved ejection fraction (HFpEF) remain undiagnosed by resting evaluation alone. Therefore, exercise testing is proposed. The diastolic stress test (DST), however, has limited sensitivity. We aimed to determine the clinical significance of adding the mean pulmonary artery pressure over cardiac output (mPAP/CO) slope to the DST in suspected HFpEF. In this prospective cohort study, consecutive patients (n=1936) with suspected HFpEF underwent exercise echocardiography with simultaneous respiratory gas analysis. These patients were stratified by exercise <i>E</i> over <i>e</i>' (exE/<i>e</i>') and mPAP/CO slope, and peak oxygen uptake, natriuretic peptides (NT-proBNP [N-terminal pro-B-type natriuretic peptide]), and score-based HFpEF likelihood were compared. Twenty-two percent of patients (n=428) had exE/<i>e</i>'<15 despite a mPAP/CO slope>3 mm Hg/L per min, 24% (n=464) had a positive DST (exE/<i>e</i>'≥15), and 54% (n=1044) had a normal DST and slope. Percentage of predicted oxygen uptake was similar in the group with exE/<i>e</i>'<15 but high mPAP/CO slope and the positive DST group (-2% [-5% to +1%]), yet worse than in those with normal DST and slope (-12% [-14% to -9%]). Patients with exE/<i>e</i>'<15 but a high slope had NT-proBNP levels and H₂FPEF (heavy, hypertensive, atrial fibrillation, pulmonary hypertension, elder; filling pressure) scores intermediate to the positive DST group and the group with both a normal DST and slope. Twenty-two percent of patients with suspected HFpEF presented with a mPAP/CO slope>3 mm Hg/L per min despite a negative DST. These patients had HFpEF characteristics and a peak oxygen uptake as low as patients with a positive DST. Therefore, an elevated mPAP/CO slope might indicate HFpEF irrespective of the DST result.