BACKGROUND: Risk models for mortality after liver transplantation have poor predictive ability. We examined whether the performance of these risk models can be improved by including information about patients' functional status (i.e., their ability to carry out activities of daily living) in addition to conventional clinical risk factors. METHODS: The UK and Ireland Liver Transplant Audit has data on all liver transplantations carried out in both countries since 1994. We examined the association of functional status measures taken immediately before transplantation on a 5-point scale (modified version of the Eastern Cooperative Oncology Group performance status) and mortality 90 days after transplantation. Logistic regression was used to adjust for other risk factors. RESULTS: Posttransplant mortality increased from 5.3% in patients able to carry out normal activity without restriction (functional status 1) to 24.8% in patients completely reliant on nursing and medical care (functional status 5; P for trend 0.003). This association remained after adjustment for conventional risk factors (adjusted P for trend 0.003). Adjusted odds ratios with functional status 3 (the most frequent functional status) as baseline category were 0.60 (95% confidence interval 0.29-1.25) for functional status 1, 0.70 (0.50-0.97) for functional status 2, 1.00 (0.71-1.41) for functional status 4, and 1.85 (1.07-3.19) for functional status 5. CONCLUSIONS: Considering a patient's functional status or more general measures of a patient's health status before transplantation in addition to conventional clinical factors may help to improve our ability to predict posttransplant survival.