Résumé : Abstract Funding Acknowledgements Type of funding sources: None. The Cardiopulmonary exercise testing (CPET) is the gold standard for cardiorespiratory and metabolic function at exercise. This CPEThighlight the limiting factors, but also provides prognostic indicators. Among the prognostic factors, maximal oxygen uptake (peak VO2) has been used for many years, particularly in heart failure, but some studies have highlighted the limitations of isolated use of peak VO2. The measurement of the Ve / VCO2 slope is currently recognized as having a complementary prognostic value. However, the manner of determining the Ve / VCO2 slope is not universally established. A single CPET could therefore deliver several slope measurements depending on the method used to calculate it. This lack of clarity would lead to possible variations affecting the reflection of the severity of the disease or the prognostic power of this parameter. Our work consisted in calculating and comparing, over nearly 700 successive CPET, the slope Ve / VCO2: 1) by calculating it up to SV1, 2) by calculating it up to SV2, 3) by the calculating on all the data beyond SV2 (in the 44% of patients in whom SV2 could be determined). The patients were classified according to the indication of the examination in 1) cardiac (54%), 2) pulmonary (15%), 3) dyspnea assessment (10%) and 4) various (oncology, obesity, diabetes, rheumatic disease: 21%). The "cardiac" category was subdivided into HTP, IC, operated CPI and non operated CPI) (results not detailed in this abstract). We have found systematic and significant differences depending on how the slope is calculated. On the whole population: in patients who did not reach SV2 (n = 389), the Ve / VCO2 slope reaches 33.8 +/- 8.5 when it is calculated up to SV1 while it reaches 38.5 + / - 8.7 (p <0.001) when it is calculated over the entire CPET. On the whole population: in patients with SV2 (n = 308), the Ve / VCO2 slope reaches 29.4 +/- 5.9 when it is calculated up to SV1, 33.5 +/- 6 (p <0.001) when it is calculated up to SV2 and 37.8 +/- 6.9 (p <0.001 and p <0.001) when it is calculated over the entire examination. When we analyze in more detail the population according to the indication of the examination, we observe differences of 3.8 to 5.2 units depending on the calculation method in patients not achieving SV2 and from 7 to 8.5 units. depending on the method of calculation in patients exceeding SV2. In subdivisions of the cardiac population the differences are also large : for example in HTP, the values ranging from 34.7 +/- 7.1 to 42.9 +/- 8.9 (p <0.001) depending on the calculation method. Our work highlights that the way of determining the Ve / VCO2 slope greatly influences its value. The significant variations observed certainly limit the prognostic value of the Ve / VCO2 slope and homogenization of the calculation method seems essential in order to improve its prognostic value.