BACKGROUND: Biliary atresia (BA) is a severe hepatobiliary disease in infants that ultimately results in hepatic failure; however, its pathological mechanism is poorly elucidated. Current surgical options, including Kasai hepatoportoenterostomy and orthotopic liver organ transplantations, are palliative; thus, innovation in BA therapy is urgent. METHODS: To examine whether BA-specific post-natal stem cells are feasible for autologous cell source for BA treatment, we isolated from human exfoliated deciduous teeth, namely BA-SHED, using a standard colony-forming unit fibroblast (CFU-F) method and compared characteristics as mesenchymal stem cells (MSCs) to healthy donor-derived control SHED, Cont-SHED. BA-SHED and Cont-SHED were intrasplenically transplanted into chronic carbon tetrachloride (CCl4)-induced liver fibrosis model mice, followed by the analysis of bile drainage function and donor integration in vivo. Immunohistochemical assay was examined for the regeneration of intrahepatic bile ducts in the recipient's liver using anti-human specific keratin 19 (KRT19) antibody. RESULTS: BA-SHED formed CFU-F, expressed MSC surface markers, and exhibited in vitro mesenchymal multipotency similar to Cont-SHED. BA-SHED showed less in vitro hepatogenic potency than Cont-SHED. Cont-SHED represented in vivo bile drainage function and KRT19-positive biliary regeneration in chronic carbon tetrachloride-induced liver fibrosis model mice. BA-SHED failed to show in vivo biliary potency and bile drainage function compared to Cont-SHED. CONCLUSION: These findings indicate that BA-SHED are not feasible source for BA treatment, because BA-SHED may epigenetically modify the underlying prenatal and perinatal BA environments. In conclusion, these findings suggest that BA-SHED-based studies may provide a platform for understanding the underlying molecular mechanisms of BA development and innovative novel modalities in BA research and treatment.