The current liver allocation system has been criticized, since available organs go to those who are the most critically ill. These recipients have the poorest overall survival. Identification of pretransplant risk factors for mortality would allow better allocation of donor livers. This study was a retrospective analysis of pretransplant clinical and laboratory parameters and subsequent postoperative liver transplant mortality to identify high-risk subgroups. Of 347 consecutive consecutive primary liver transplant recipients, 59 (17%) met United Network for Organ Sharing (UNOS) criteria for status 4. Pretransplant factors included liver function, coagulation, albumin and ammonia levels, renal function, the presence of ascites, and etiology of liver disease. Overall 1-year patient survival was significantly worse for the status 4 recipients (89.0% vs. 67.7%; P=0.01). In a univariate analysis of pretransplant risk factors for all recipients, elevated creatinine (P=0.008) and ammonia (P=0.009), and UNOS status 4 (P=0.01) significantly affected postoperative survival. In multivariate analysis of pretransplant risk factors for all recipients, elevated creatinine (P=0.003) was the only factor to significantly affect postoperative survival. In UNOS status 4 patients, univariate analysis of pretransplant risk factors and their influence on patient survival demonstrated that prolonged coagulation partial thromboplastin time (P=0.04) and a higher grade of encephalopathy (P=0.02) significantly affected postoperative survival. Advanced encephalopathy (P=0.009) and prolonged partial thromboplastin time (P=0.01) were the only significant risk factors by multivariate analysis in status 4 patients. In status 4 and non-status 4 patients, we identified risk factors that adversely affected patient survival, but their predictive power was insufficient to deny transplantation. Despite the higher mortality in status 4 recipients, their long-term survival is only slightly worse than that of non-status 4 patients. Until better predictors of survival are ascertained, our data do not support limiting the use of donor livers in UNOS status 4 recipients. © 1995 by Williams and Wilkins.