For ideal classification of chronic disease in large populations, identification of the presence or absence of every chronic disease in every person would be desirable, as would knowledge of specific causes and pathogenesis, means to define cause, measure severity, and identify important milestones. The authors have approached the problems by analyzing advantages and disadvantages of certain aids to classification of chronic disease, namely, categorical disease classification, disability classification, and multiphasic screening. They describe a modified approach to classification, developed in conjunction with studies of the chronically ill and aged, which consists of establishing for each person a profile that includes principal diagnosis, disease status defined by chronic disease screening, and disability status based on comparable and systematic evaluation. A set of classes which combines the advantages of each of the separate approaches to chronic disease description is synthesized. An illustration of the use of this approach in describing the chronic disease status of one group is presented. The approach can be adapted to other populations and other purposes, and the authors suggest that it should be of particular use in large-scale population studies.