Ischemic heart disease leading to heart failure (HF) portends a high overall morbidity and mortality. A higher N-terminal-pro-B-type natriuretic peptide (NT-pro-BNP) at rest reflects HF severity and impaired cardiac output, most often secondary to reduced ejection fraction (EF). As an insufficient increase in cardiac output during exertion is common in all HF phenotypes, we examined the value of NT-pro-BNP during exercise testing as a risk stratification index for ischemic HF secondary to either reduced (HFrEF) or mid-ranged/preserved EF (HFmrEF/HFpEF). 213 patients (123 HFrEF; 90 HFmrEF/HFpEF) underwent cardiopulmonary exercise testing (CPET). NT-pro-BNP was determined at rest and peak exercise. The distribution of HFrEF and HFmrEF/HFpEF etiology in subjects with and without oxygen consumption trajectory flattening during CPET was similar (p > 0.05). Patients with HFrEF had higher plasma levels of NT-pro-BNP at rest and peak exercise than those with HFmrEF/HFpEF (984 vs. 780; 1012 vs. 845 pg/mL, p < 0.01, respectively), whereas ΔNT-pro-BNPpeak/rest was similar (60 vs. 50 pg/mL, p > 0.05). During the tracking period (22.4 ± 20.3 months) 34 patients died, and there were 2 cardiac transplantations and 3 LVAD implantations. In a multivariate regression model only the NT-pro-BNPpeak and ΔNT-pro-BNPpeak/rest were retained in the regression for the prediction of adverse events (Chi-square:8.97, p = 0.003). ROC analysis demonstrated that NT-pro-BNPpeak ≥1506 pg/mL and ΔNT-pro-BNPpeak/rest ≥108 pg/mL were optimal for identifying patients with a risk (Sn = 76.9, 74.4 %; Sp = 84.7, 80.9 %, respectively). NT-pro-BNP changes during effort and absolute peak values reached provide novel insights emerging as new and strong predictors of adverse events in HF of any EF.