Background and Aims: Survival of patients with end-stage liver disease is variable and difficult to predict. A two-phase prospective cohort study was conducted at five teaching hospitals to develop and evaluate a model for prediction of death. Methods: Five hundred thirty-eight hospitalized patients with a history of chronic liver disease and two or more signs of decompensation were studied. Results: The cumulative incidence of death was 30% at 30 days and 50% at 6 months. In 295 patients in phase I, time till death was independently associated (P < 0.01) with five factors measured on study day 3: renal insufficiency, cognitive dysfunction, ventilatory insufficiency, age ≤ 65 years, and prothrombin time ≤16 seconds. These risk factors stratified 243 patients in phase II into three groups with cumulative incidences of death at 30 days of 12%, 40%, and 74%, respectively. Integration of the prognostic model with physicians' predictions led to improved estimates of the probability of death. Although performance of liver transplantation after study entry was independently associated with enhanced survival, the intensity of other acute therapies was not. Conclusions: Five risk factors were associated with the risk of death in patients with end- stage liver disease and provided a quantitative basis to complement physicians' prognostic estimates.