Objective To compare the advantages of critically ill neonates mortality risk prediction between neonatal critical illness score (NCIS) and pediatric risk of mortality score and to research for the better neonatal disease severity scoring system feasible for use in our country.Methods The 135 neonates, who were in neonatal intensive care unit（NICU）from January 2012 to June 2013 in the first Hospital Affiliated to Shihezi University, were scored by NCIS and pediatric risk of mortalityⅢ (PRISM Ⅲ).According to the simultaneously scores the patients were divided into three groups（very critical group, critical group, non-critical group）, and mortality rates were compared in each group.At the same time the receiver operating characteristic curve（ROC） was drawn and counted out their areas under the ROC curves (AUC) to observe the specificity and sensitivity of two scoring systems in predicting the risk of death in critically ill neonates.Results The comparison between NCIS and PRISM III score:the differences were statistically significant in mortality among the noncritical, the critical and the very critical groups（P all ＜0.05）;the differences were not statistically significant in mortality among the non-critical, the critical and the very critical subgroups（P all ＞0.05）;AUC:NCIS 0.900, PRISM III 0.909.Conclusion Both NCIS and PRISM III can predict mortality risk in critically ill neonates, PRISM III is fully applicable in China.