Abstract Background Cardiopulmonary exercise testing is a non-invasive alternative in the evaluation of subjects with heart failure and because it comprehensively evaluates the respiratory and cardiovascular systems under conditions of physical stress. The objective is validity of the values in the ventilatory equivalents in the diagnosis of heart failure with reduced left ventricular ejection fraction. Methods Cross-sectional study in subjects with cardiopulmonary exercise testing and transthoracic echocardiography. The participants had a diagnosis of heart failure according to the American Heart Association, a group with reduced left ventricular ejection fraction < 40% and no reduced ≥ 40%, this last group included patients with intermediate or normal ejection fraction. Results In the final analysis, 138 patients were included, of whom 88% (122/138) had reduced left ventricular ejection fraction and 12% (16/138) not reduce, mean age was 58.9 years (SD: 16.37), 65% were men, and the body mass index was 27.6 (SD: 4.56). In patients with reduced left ventricular ejection fraction, maximum oxygen consumption presented an average of 1.6 l/min (SD: 0.68) compared to 1.7 l/min (SD: 0.66) in the control group (p = 0.513). The ventilatory equivalents for carbon dioxide (VE/VCO2) during the ventilatory and maximum threshold presented an average of 38.4 (SD: 7.38) and 44.6 (SD: 8.24), respectively. The discriminatory performance of VE/VCO2 versus peak oxygen consumption was 0.737 (95% CI: 0.596–0.878; p = 0.008) in patients with left ventricular ejection fraction < 40%. Conclusions VE/VCO2 was a reliable predictor of normal or reduced left ventricular ejection fraction and performs well against peak oxygen consumption.