Objective: To determine the accuracy of the Pulse Oximetry Test (POT) in screening for Congenital Heart Diseases (CHD) in newborns in the first 48 hours of life. Method: Systematic review of diagnostic test accuracy with meta-analysis. The selection of studies was carried out in June 2021. Studies were selected with newborns, in a hospital or home environment, without a previous diagnosis of CHD, regardless of gestational age at birth, who underwent POT within the first 48 hours after birth. Registration on the PROSPERO platform – CRD42021256286. Results: Twenty-nine studies were included, totaling a population of 388,491 newborns. POT demonstrated sensitivity of 47% (95% CI: 43% to 50%) and specificity of 98% (95% CI: 98% to 98%). Subgroup analyses were carried out according to the different testing period, inclusion of retests in protocols and population of premature newborns. Conclusion: POT is a test with moderate sensitivity and high specificity. It is more effective when carried out within 24h – 48h of birth; in protocols that present retests, within two hours after the first measurement. It does not show satisfactory effectiveness for premature newborns.