PURPOSEThis study aimed to investigate baseline, exercise testing, and exercise training–mediated predictors of change in peak oxygen uptake (V˙O2peak) from baseline to 12-wk follow-up (ΔV˙O2peak) in a post hoc analysis from the SMARTEX Heart Failure trial. METHODSWe studied 215 patients with heart failure with left ventricular ejection fraction (LVEF) ≤35%, and New York Heart Association (NYHA) classes II–III who were randomized to either supervised high-intensity interval training with exercise target intensity of 90%–95% of peak heart rate (HRpeak) or supervised moderate continuous training (MCT) with target intensity of 60%–70% of HRpeak, or who received a recommendation of regular exercise on their own. Predictors of ΔV˙O2peak were assessed in two modelsa logistic regression model comparing highest and lowest tertiles (baseline parameters) and a multivariate linear regression model (test/training/clinical parameters). RESULTSThe change in V˙O2peak in response to the interventions (ΔV˙O2peak) varied substantially, from −8.50 to +11.30 mL·kg·min. Baseline NYHA (class II gave higher odds vs III; odds ratio (OR), 7.1 (2.0–24.9); P = 0.002), LVEF (OR per percent, 1.1 (1.0–1.2); P = 0.005), and age (OR per 10 yr, 0.5 (0.3–0.8); P = 0.003) were associated with ΔV˙O2peak.In the multivariate linear regression, 34% of the variability in ΔV˙O2peak was explained by the increase in exercise training workload, ΔHRpeak between baseline and 12-wk posttesting, age, and ever having smoked. CONCLUSIONExercise training response (ΔV˙O2peak) correlated negatively with age, LVEF, and NYHA class. The ability to increase workload during the training period and increased ΔHRpeak between baseline and the 12-wk test were associated with a positive outcome.