Postcardiotomy cardiogenic shock (PCCS) is a rare clinical entity associated with substantial morbidity and mortality. It is characterized by heart failure that results in an inability to be weaned from cardiopulmonary bypass (CPB). The aim of this study was to analyze the outcomes of extracorporeal membrane oxygenation (ECMO) in patients with PCCS and to identify predictors of in-hospital mortality and failure to be weaned from ECMO. From January 2002 to August 2016, 3248 patients underwent cardiac surgery in our hospital. Of these, 29 patients (0.89%) required ECMO because of an inability to be weaned from cardiopulmonary bypass. The median duration of ECMO support was 144 h (340–52 h) (range 17–818 h). Sixteen patients (55.2%) were weaned from ECMO, and 6 (20.7%) survived to hospital discharge. The multivariate analysis revealed that reoperation [odds ratio (OR): 13.667, 95% confidence interval (CI): 0.999–187.056, p = 0.05] and ECMO support duration > 130 h (OR: 17.688, 95% CI: 1.324–236.233, p = 0.03) were independent predictors of failure to be weaned from ECMO. Temporarily being weaned from CPB > 15 min (OR: 0.027, 95% CI: 0.001–0.586, p = 0.02) was found to be a protective factor. The multivariate analysis revealed that CPB time > 270 min (OR: 12.503, 95% CI: 1.058–147.718, p = 0.05) and ECMO support duration > 60 h (OR: 12.503, 95% CI: 1.058–147.718, p = 0.05) were independent predictors of in-hospital mortality. ECMO is an acceptable technique for treating PCCS in patients undergoing cardiac surgery. Our data suggest a reevaluation of therapeutic strategies after 60 h and again after 130 h of ECMO support.