BACKGROUND & AIMS: Liver disease in Alagille syndrome is highly variable. Many of the patients presenting with severe cholestasis early in life improve spontaneously; 10-20%, however, have progressive disease. It is currently not possible to predict long-term hepatic outcomes in Alagille syndrome. This international, multicentre study was aimed at identifying early life predictors of liver disease outcome. METHODS: Retrospective clinical, laboratory and radiographic data from a cohort of 144 Alagille syndrome patients, whose long-term hepatic outcomes had been determined a priori based on previously published criteria, were collected. RESULTS: Sixty-seven patients had mild and 77 had severe hepatic outcome. Univariate analysis demonstrated that cholestasis and fibrosis on biopsy, as well as the presence of xanthomata were significantly different between the groups (P < 0.05 for all). Mixed model analysis revealed that total serum bilirubin and serum cholesterol were also associated with outcome (P = 0.001 and P = 0.002, respectively). Graphical representation of the data revealed a change in total bilirubin levels between 12 and 24 months of age in the mild group. Recursive partitioning identified a threshold for total bilirubin of 3.8 mg/dl (65 mmol/L) in that age-frame that differentiated between outcomes. A multivariable logistic regression model was developed using fibrosis, xanthomata and the total bilirubin cut-off of 3.8 mg/dl (65 mmol/L), which generated an area under the ROC curve of 0.792. CONCLUSIONS: The long-term hepatic outcomes of patients with Alagille syndrome can be predicted based on serum total bilirubin between the ages of 12-24 months combined with fibrosis on liver biopsy and the presence of xanthomata on physical examination.