Abstract Introduction One crucial element in cardiopulmonary exercise testing (CPET) involves identifying the first and second ventilatory thresholds (VT), VT1 and VT2, which hold known prognostic implications. If the individual can sustain exercise considerably beyond VT2, a third breakpoint (VT3 – Figure A) may emerge, coinciding with heightened hyperventilation to counter metabolic acidosis. Historically, this point has been mainly described among athletes and its meaning is still not entirely clear. We aimed to assess the prognostic value of VT3 in patients with heart failure (HF). Methods Retrospective, single-center study enrolling HF patients undergoing CPET from 2015 to 2021. We evaluated the typical CPET parameters along with the three ventilatory thresholds: VT1, VT2, and the newly introduced VT3. The primary outcome was a composite of cardiovascular death, urgent heart transplant, need for left ventricle assistance device and HF hospitalizations. Results We included 221 patients (82% men, mean age 58±12 years), among whom 69% had ischemic heart failure aetiology. The mean left ventricular ejection fraction (LVEF) was 34±9%, with a mean respiratory exchange ratio 1.15±0.082, mean pVO2 18.5±6.3, and a mean VE/VCO2 slope of 41.3±12.9. VT3 was attained by 43% (n = 94) of the patients. In comparison to those who did not achieve VT3, patients who reached this threshold were younger (56±11 vs 60±12; p = 0.05), had a higher pVO2 (20.5±6.9 vs 17±5.4; p = 0.05), higher RER (1.16±0.86 vs 1.13±0.78; p = 0.05), and lower VE/VCO2 slope (39±10.3 vs 43±14.3; p = 0.05). Regarding the three ventilatory thresholds (VT1, VT2, VT3), 7% (n=17) achieved a maximum of one threshold, 50% (n=110) achieved two thresholds, and 43% (n=94) reached all three thresholds. After a median follow-up of 2,3 years patients with identifiable VT3 had significantly better prognosis (HR 0,434; 95%CI 0,246-0,767; p=0,004) than those without this threshold. Furthermore, the number of identifiable ventilatory thresholds correlated positively with prognosis. In fact, patients with only 2 thresholds have twice the probability of developing an event and those with one or zero identifiable thresholds have a 4-fold increase in the probability of events (Figure C). Conclusion In a cohort of patients with HF, those with identifiable VT3 had a lower incidence of events. This parameter may aid in prognostic stratification in this population. However, this findings warrant further prospective validation.Figure 1