To systematically examine the relative magnitude and predictors of responses to exercise training in older adult with heart failure (HF) with reduced ejection fraction (HFrEF), and preserved EF (HFpEF). Secondary analysis of a randomized controlled trial. Outpatient cardiac rehabilitation program. Individuals with HF (24 HFrEF, 24 HFpEF) who underwent supervised exercise training. The study included individual-level data from the exercise training arms of a randomized controlled trial that evaluated the effect of 16weeks of supervised moderate-intensity endurance exercise training in older adults with chronic, stable HFpEF and HFrEF. Changes in peak oxygen uptake (VO2peak ) in response to supervised training in individuals with HFpEF were compared with that of individuals with HFrEF. The significant clinical predictors of changes in VO2peak with exercise training were assessed using univariate and multivariate regression models. Training-related improvement in VO2peak was higher in participants with HFpEF than in those with HFrEF (change: 18.7±17.6% vs -0.3±15.4%, P<.001). In univariate analysis, echocardiographic abnormalities in left ventricular structure and function and lower body mass index were associated with blunted response of VO2peak with exercise training. In multivariate regression analysis using stepwise selection, submaximal exercise systolic blood pressure, and resting early deceleration time were independent predictors of change in VO2peak . The change in VO2peak in response to endurance exercise training in older adults with HF differssignificantly according to HF subtype, with greater VO2peak improvement in HFpEF than HFrEF. These results suggest that the current Centers for Medicare and Medicaid Services policy excluding individuals with HFpEF from reimbursement from cardiac rehabilitation may need to be revisited.