This paper presents various laboratory methods designed to estimate body composition and documents some of the progress in their validation. In addition to the well-known methods of densitometry, hydrometry, and spectrometry (40K), many other methods will be reviewed briefly for their relation to body composition. Emphasis is given to validation principles which need to be followed if new methods are to be developed. The reliance of past research on the two-component model and reference man is reviewed, and the need for multicomponent approaches to the study of body composition is emphasized. Past research using the two-component approach has until recently led to lack of research in developing new methodologies, has limited the potential usefulness of various laboratory methods in estimating body composition in different populations, and has made the relation of body composition to health, performance, and exercise an inexact science. The application of several multicomponent approaches to the characterization of both fat-free body composition and body composition of various populations will lead to the development of reference bodies so essential for the advancement of the field. The estimation of body composition changes with exercise, growth, development, and aging and the relation of body composition to health and physical performance are important areas for future research using various multicomponent approaches.