Objective To investigate the accuracy of pulse oximetry screening for congenital heart disease （CHD） in asymptomatic children. Methods We electronically searched VIP database，CNKI database，CBMdisc，Wanfandatabase，PubMed，Excerpta Medica Database （EMBASE） and the Cochrane Library from the time of database establishment to February 2013 to identify literatures regarding pulse oximetry screening for CHD in asymptomatic children. Data extraction was performed by two researchers independently. Quality of the included literatures was evaluated with qualityassessment for diagnostic accuracy studies （QUADAS）. Meta-analysis was performed using RevMan 5.1.1. ResultsTwelve studies were included in this study with a total of 137 582 newborns. Fourteen QUADAS criteria were fulfilled by all the 12 studies. Among the 12 studies，only the studies with positive test results received verification using a gold standardtest，and the execution of the gold standard test was described in sufficient detail only in 1 study. There were 10 studies in which the gold standard test results were interpreted with knowledge of the diagnostic test results. Meta-analysis showed that pooled sensitivity and specificity of pulse oximetry test for the diagnosis of CHD were 22% with 95% CI （19%，25%） and 99% （99%，99%） respectively. Pooled positive likelihood ratio （LR＋） and negative likelihood ratio （LR-） were 157.30 with 95% CI （11.80，2 096.95） and 0.61 with 95% CI （0.46，0.82） respectively. Pooled diagnostic odds ratio was 398.25 （34.5，4 596.81）. The area under summary receiver operating characteristic （SROC） curve was 0.809，and Q index was 0.744. Conclusion Pulse oximetry is a lowly sensitive and highly specific diagnostic method for children with CHD，and is helpful for early diagnosis of CHD.