Background: Patients with suspected heart failure and preserved ejection fraction (HFpEF) undergo invasive (right heart catheterization) cardiopulmonary exercise testing (iCPET) to confirm diagnosis. During iCPET peripheral arterial blood pressure (BP) is measured using a brachial cuff or radial arterial catheter. Notably, peripheral BP is typically higher than central BP and this difference is assessed by measuring pulse pressure amplification (PPA; peripheral - central PP; PP = systolic BP [SBP] - diastolic BP). The effect of PPA on the assessment of (exaggerated) exercise BP in HFpEF is unknown. Methods: 20 patients with HFpEF (75% Female; age: 69±7 years, BMI: 38±7 kg/m 2 ) were studied at rest and during submaximal and maximal cycling exercise. We measured radial BP (arterial catheterization) and derived ascending aortic BP using a transfer function applied to radial arterial waveforms (SphygmoCor). Pulse pressure amplification was assessed as radial-aortic pulse pressure. Data were compared with one-way and two-way (Site*Exercise Stage) ANOVA. Results: The increase in radial PP with exercise was greater than aortic PP (Site*Stage P<0.001; Figure A); thus, PPA increased with exercise (ANOVA, P =0.001; Figure B). The increase in PPA was due to a greater increase in radial compared to aortic SBP (Site*Stage P=0.002; Figure C). The change in diastolic pressure was not different between sites (Figure D). Increases in PPA accounted for ~30% of the increase in radial SBP. The incidence of exaggerated exercise SBP (Male: >210 mmHg; Female: >190 mmHg) was higher in the radial artery compared to aorta (95 vs 40 %, χ2<0.001). Conclusion: Peak exercise SBP was ~50 mmHg greater in the radial artery compared to aorta. The marked overestimation of central BP during exercise when measuring peripheral BP leads to misguided interpretation of BP responses to exercise, the magnitude of afterload and has implications for the assessment of ventricular-arterial coupling.