Abstract Background All the studies and meta-analyses showed beneficial effects of the exercise based cardiac rehabilitation (ExBCR) on peak value of oxygen uptake (VO2peak) in patient with chronic heart failure (CHF). It is widely recognized that changes in VO2peak induced by training (ΔVO2peak) have a prognostic value. A 5.0% increase in ΔVO2peak has been associated with a significant clinical improvement1. However, 20 to 30 % of CHF are non-responders to ExBCR. The Signal-Morphology Impedance Cardiography (SM-ICGTM) technology has allowed previous researchers to show the role of cardiac output (CO) on DVO2peak2. CO responses may be partly attributed to the peak ejection flow velocity that can explain the impaired response to ExBCR. Purpose The study aimed to establish whether SM-ICGTM is an appropriate technology to predict responses to ExBCR in CHF patients. Methods A retrospective study was conducted on 65 CHF patients (LVEF: 29.0±8.3%). All patients performed a cardiopulmonary exercise test (CPET) on bicycle to determine ventilation (VE) and VO2peak before and after 20 ExBCR sessions. In addition, stroke volume (SV), heart rate (HR), and CO were continuously monitored by SM-ICGTM (PhysioFlow® Lab1). The value of peak ejection flow velocity, so called CTi by the manufacturer, was calculated as the peak of the first mathematical derivative over the time of the impedance waveform during the systole and measured by PhysioFlow® throughout the test. For the analysis of training response patients were divided in two groups: the normal CTi profile which was characterized by an initial increase followed by a plateau until CPET exhaustion, and the abnormal CTi response if CTi decreased at the onset of CPET or increased then decreased before exhaustion. Statistical analyses were performed with JASP. The significant level was set at p value < 0.05 for 95% confidence interval. Results Our study population consisted in 32 patients who presented a normal CTi profile and 33 with an abnormal CTi profile. As shown in Table 1, no significant difference was found at baseline between both groups. ExBCR induced significant changes in VO2peak even if 29 % patients were non-responders. Patients with normal CTi profile presented a greater DVO2peak compared to patients with abnormal CTi profile (28 ± 30 vs. 14 ± 19 %, p<0.01). 78% of patients with normal CTi profile versus only 64% of patients with abnormal CTi profile were ExBCR responders. The logistic regression showed that patient with normal CTi profile were 3.3 time more likely to present a ΔVO2peak ≥ 5%. Conclusion Exercise CTi profiles determined by SM-ICGTM prove to be an interesting tool to help predicting ExBCR responses in CHF patients and maybe highlighting the hemodynamic determinants of VO2 responses to exercise training.