Although a variety of techniques are currently available to measure body composition in human beings, each has limitations with respect to its applicability in addressing nutrition research issues. Several methods "directly" determine an individual's body fat or lean tissue mass with sufficient precision and accuracy such that the data are appropriate for making fine distinctions between individuals or within an individual over time. Such direct methods are most useful in nutritional/metabolic research that requires such distinctions. Each of these techniques will be discussed. There also exist a number of "indirect" measures of body composition. These techniques require more interpretative assumptions and are perhaps better suited to population studies, in which it is often less critical to make fine distinctions between individuals. The appropriate design and interpretation of many nutrition-related research studies depend upon the correct use of body composition data. This review will attempt to help the reader understand the principles of each of the currently available body composition techniques and its optimal use.