We aimed to analyze the outcome and identify predictors of hospital mortality in patients with refractory cardiac arrest (CA) complicating acute coronary syndromes (ACS) and requiring veno-arterial extracorporeal membrane oxygenation (VA-ECMO) treatment. Between Jan-2005 and Dec-2019, 51 patients underwent urgent VA-ECMO implantation for CA in ACS. Patients were divided in two groups: "in-hospital" cardiac arrest (IHCA) and "out-of-hospital" cardiac arrest (OHCA). Prospectively collected data were retrospectively analyzed and compared between groups. Predictors for hospital mortality were investigated. IHCA and OHCA patients were 32 (62.7%) and 19 (37.3%), respectively. The groups differed for: male gender (72% vs 95%; p = 0.070), lactate peak level (8.5 ± 4.3vs10.7 ± 2.9; p = 0.023), total elapsed time from CA to VA-ECMO implantation in both groups (p < 0.001) and elapsed time from CA (IHCA group) or hospital arrival (OHCA group) to VA-ECMO implantation (38 min vs 80 min; p = 0.001). At logistic regression analysis, concomitant lactate level greater than 8.0 mmol/L and elapsed time from CA to VA-ECMO ≥ 30 min were predictors of increased mortality (OR 3.9; 95% CI 1.19-12.79; p = 0.025) for the entire population. In-hospital mortality was 60.8% (31/51 patients): 68.4% in OHCA group and 56.2% in IHCA group. No risk factors related to 30-day mortality resulted significant at univariable analysis. When rapidly instituted, VA-ECMO improves survival in patients with refractory cardiac arrest allowing coronary syndrome treatment. The association of an elapsed time from CA to VA-ECMO implantation longer than 30 min and a preoperative lactate peak level over 8.0 mmol/L predict a poor outcome, independently from being IHCA or OHCA.