Background: Early distinguishing biliary atresia from other causes of infantile cholestasis remains a major challenge. We aimed to develop and validate a scoring system based on bile acid for identification of biliary atresia.Methods: In a prospective study, a total of 141 infants with cholestasis were enrolled in two sets (derivation cohort, n=66; validation cohort, n=75) from 2014 to 2018. Variables with significant difference between biliary atresia and non-biliary atresia infants were selected in the derivation cohort. Then, a scoring system including those variables was designed and validated. Results: Among 66 patients in the derivation cohort, 34 (51.5%) had biliary atresia. A scoring system was proposed with the following variables: glycochenodeoxycholic acid/chenodeoxycholic acid, clay stool, and gamma-glutamyl transferase. The total score ranged from 0 to 41, and a cutoff value of 15 identified biliary atresia with an area under receiver operating characteristic curve of 0.87 (95% confidence interval, 0.77–0.94), sensitivity of 85.3%, and specificity of 81.3% in the derivation cohort; these values were also confirmed in a validation cohort with a sensitivity of 90.0% and specificity of 80.0%. Conclusions: The proposed simple scoring system had good diagnostic accuracy for estimating the risk of biliary atresia in infants with cholestasis.