Introduction: Right ventricular dysfunction (RVD) is a marker of adverse outcome in HFpEF patients. While RVD in HFpEF has been demonstrated using dobutamine stress and invasive hemodynamic studies, detailed assessment of RV and pulmonary vascular (PV) function using simultaneous exercise cardiac MRI and invasive hemodynamic measurements has never been performed. Methods In a prospective study, we performed CPET to determine maximal power output (Pmax) in patients with HFpEF (n=11, age 73±9 years) and in healthy controls (n=8, age 53±7 years). We also investigated asymptomatic subjects with echocardiographic evidence of LV concentric remodeling (LVCR) and impaired relaxation (n=4, age 63±7 years). All participants then performed a three-stage supine bicycle exercise test during real-time CMR imaging (at 25%, 50% and 66% of Pmax) with continuous recording of pulmonary arterial pressures using fluid-filled catheters. We determined the ratio of mPAP relative to cardiac output (mPAP/CO slope) as a measure of PV reserve and the change in RV end-systolic pressure volume ratio (dRVESPR) as a measure for RV contractile reserve. Results At rest, there was no difference in LV or RV ejection fraction (EF) between HFpEF patients and controls (p=NS). However, during exercise, HFpEF patients showed a decline in RVEF (-3.9±1.7%), contrasting with a marked increase in RVEF in the control group (+12.49±2.33%, p<0.001 for interaction workload*group, figure 1A). HFpEF patients had a greater mPAP/CO slope (p<0.001, figure 1B) and a significantly lower dRVESPVR (p<0.001) than controls. Interestingly, asymptomatic subjects with LVCR showed an intermediate response to exercise (figure 1A-B). Conclusion In patients with HFpEF, RV functional reserve and pulmonary pressure-flow relationships are markedly impaired. These data suggest an important role for cardiac exercise MRI in phenotyping heart failure patients and guiding RV or pulmonary vasculature targeted therapy.