Objectives: We examined a large single-institution experience in extracorporeal cardiopulmonary resuscitation (ECPR) in children having cardiac arrest refractory to conventional resuscitation measures with focus on factors affecting survival. Methods: Between 2002 and 2017, 184 children underwent ECPR at our institution. We entered demographic, anatomic, clinical, surgical, and ECPR support details into a multivariable logistic regression models to determine factors associated with mortality. Results: Median age was 54 days (interquartile range [IQR]: 11-272). In all, 157 (85%) patients had primary cardiac disease, including 136 (74%) with congenital heart disease (71 with single ventricle). Extracorporeal cardiopulmonary resuscitation occurred following cardiac surgery in 124 (67%) patients. Median cardiopulmonary resuscitation (CPR) duration was 27 minutes (IQR: 18-40) and median support duration was 3.0 days (IQR: 1.6-5.3). Overall, ECPR was weaned in 115 (63%), with 79 (43%) surviving to hospital discharge. Survival for patients with congenital heart disease, noncongenital cardiac, and noncardiac pathologies was 44%, 71%, and 15%, respectively. On multivariable regression analysis, risk factors associated with mortality were presupport pH <7.1 (odds ratio [OR] = 3.7, 95% confidence interval [CI]: 1.11-12.41, P = .033), mechanical complications (OR = 8.33, 95% CI: 1.91-36.25, P = .005), neurologic complications (OR = 6.27, 95% CI: 1.40-28.10, P = .017), and renal replacement therapy (OR = 3.31, 95% CI: 1.03-10.66, P = .045). Conclusions: Extracorporeal cardiopulmonary resuscitation plays a valuable role salvaging children with refractory cardiac arrest. Survival varies with underlying pathology and can be expected even with relatively longer CPR durations. Efforts to improve systemic output before and after institution of ECPR might mitigate some of the significant risk factors for mortality.