Aims & Objectives: Extracorporeal membrane oxygenation (ECMO) is a rescue therapy for refractory respiratory or cardiovascular failure. In children, extremely low tidal volumes have been associated with increased mortality in PARDS patients on ECMO. The effect of ultra-lung-protective ventilation on clinical outcomes of interest is not clear. Methods: A retrospective cohort study was carried out (January 2016-January 2024) in a university hospital. Mechanical ventilation parameters were evaluated during the first 3 days of ECMO support. The primary outcome was the association with mortality between lung-protective ventilation (LPV) defined as Vt 4-6 mL/kg with driving pressure (DP) less than 15 cm/H20 compared with ultra-lung-protective ventilation (ULPV) defined as Vt < 4mL/kg with DP less than 15 cm/H20. Results: A total of 102 children (age of 1 year [IQR 1-9]) were included. Patients with extremely low Vt (less than 4 mL/kg) have higher odds of mortality regardless of age and PEEP level (aOR 2.56 95% CI 1.12-5.87; p=0.02) when the DP is > 15 cm/H20. Children with LPV have lower odds of mortality (aOR 0.41 95% CI 0.17-0.99; p=0.04) compared to ULPV regardless of the age and PEEP level. We did not find a higher frequency of atelectasis (p=0.24) or a longer duration of hospital stay after decannulation (p=0.32) between the two groups. Conclusions: In children with ECMO support in LPV we found lower odds of mortality compared to ULPV without differences in the frequency of atelectasis or length of hospital stay. Clinical trials are needed to help confirm these findings. Keywords: Hypoxia, ECMO, Mortality, PARDS, respiratory failure