Albumin–bilirubin (ALBI) grade is defined using the ALBI score, which is calculated based on total serum bilirubin and albumin. This study aimed to evaluate the diagnostic ability of the ALBI score for determining hepatic fibrosis stage and transplant-free survival in primary biliary cholangitis (PBC) patients. A total of 181 Japanese patients with biopsy-proven or serologically diagnosed PBC were enrolled. The pathological stage was assessed using the Scheuer classification. The ALBI score differentiated fibrosis in stage 4 from that of 3 in the biopsy-proven cohort (p < 0.05). With an ALBI score cut-off value of −1.679, the sensitivity and specificity were 100% and 91.1%, respectively, with a likelihood ratio of 12.3 to differentiate stage 4 from stages 1–3. The ALBI score at the beginning of ursodeoxycholic acid (UDCA) prescription correlated with the two prognostic scores calculated after 1-year UDCA treatment. Kaplan–Meier analysis showed that the baseline ALBI score differentiated liver transplant-free survival (p < 0.05). The ALBI score presented a greater hazard ratio for transplant-free survival than aspartate aminotransferase-to-platelet ratio index (APRI) in Cox proportional hazard model. In conclusion, ALBI score indicates pathological stage in Japanese PBC patients and scores before UDCA prescription predict better liver transplant-free survival, which correlated well with the two major prognostic scores. The prognosis-predicting ability of the ALBI score might surpass that of APRI.