A framework for evaluating and conducting prognostic studies is proposed. Recently published studies on prognosis of cirrhosis of the liver are evaluated according to the proposed framework. It was found that appropriate statistical techniques were often used to analyze the results of prognostic studies of cirrhosis of the liver. On the other hand, the studies performed poorly with regards to study design, the determination of the usefulness of the data, the validity of the collected information, and the analytic strategy. It is hoped that the criteria suggested in this paper will improve the planning and the reporting of prognostic studies.