Background: Equations have been proposed by the Wilford Hall/Santa Rosa (WHSR) and Congenital Diaphragmatic Hernia Study Group (CDHSG) for predicting survival in patients with CDH. The CDHSG stratifies risk based on a logistic regression equation incorporating birth weight and 5-min Apgar score, while the WHSR group uses the difference between maximum pO2 and maximum pCO2 as an index of risk. These models have not been applied specifically to the CDH ECMO (extracorporeal membrane oxygenation) population, a group at highest mortality risk. Objectives: To evaluate the WHSR and CDHSG predictive equations when applied to a population of patients with CDH requiring ECMO life support. Methods: A single-center retrospective review was conducted on infants with CDH treated with ECMO between 1993 and 2007. Predicted and actual outcomes were compared using receiver operating curve (ROC) analyses in which an area under the curve (AUC) of 1 denotes 100% agreement between predicted and actual outcomes. Kaplan-Meier analyses were also used to compare survival of patients who were risk-categorized according to each prediction model. Minimum pre-ECMO pCO2 was likewise evaluated as a predictor of survival. Results: Overall survival was 50% in 62 CDH patients treated with ECMO during the study period. The CDHSG equation did not discriminate between survivors and nonsurvivors (AUC 0.55, p = 0.499). The modified WHSR formula showed better discrimination of survival (AUC 0.71, p = 0.004). Lowest achievable pre-ECMO pCO2 had the highest AUC (0.723, p = 0.003). Patients with minimum pre-ECMO pCO2 <50 mm Hg had 56% survival, while those with >70 mm Hg had 0% survival. Conclusions: Equations proposed to predict survival in CDH patients may not discriminate survivors from nonsurvivors in the ECMO population. In this highest risk group, factors such as birth weight and Apgar score are less critical in estimating mortality risk than indicators of ventilation and oxygenation that reflect the degree of pulmonary hypoplasia.