Résumé : Cardiopulmonary exercise testing (CPET) was limited to peak oxygen consumption analysis (VO2peak), and now the ventilation/carbon dioxide production (VE/VCO2) slope is recognized as having independent prognostic value. Unlike VO2peak, the VE/VCO2 slope does not require maximal effort, making it more feasible. There is no consensus on how to measure the VE/VCO2 slope; therefore, we assessed whether different methods affect its value. This is a retrospective study assessing sociodemographic data, left ventricular ejection fraction, CPET parameters, and indications of patients referred for CPET. The VE/VCO2 slope was measured to the first ventilatory threshold (VT1-slope), secondary threshold (VT2-slope), and included all test data (full-slope). Of the 697 CPETs analyzed, 308 reached VT2. All VE/VCO2 slopes increased with age, regardless of test indications. In patients not reaching VT2, the VT1-slope was 32 vs. 36 (p < 0.001) for the full-slope; in those surpassing VT2, the VT1-slope was 29 vs. 33 (p < 0.001) for the VT2-slope and 37 (all p < 0.001) for the full-slope. The mean difference between the submaximal and full-slopes was ±4 units, sufficient to reclassify patients from low to high risk for heart failure or pulmonary hypertension. We conclude that the method used for determining the VE/VCO2 slope greatly influences the result, the significant variations limiting its prognostic value. The calculation method must be standardized to improve its prognostic value.