Background:Loss of lean‐body mass has been found to be predictive of death from wasting in HIV‐infected individuals. Several clinically applicable, noninvasive methods for estimating body wasting are available, but the comparability of these methods is not known. The objective of this study was to assess the agreement between estimates of lean‐body mass in HIV‐infected men.Methods: Lean‐body mass was measured by bioelectrical impedance assessment, by prediction equations that used anthropometric measurements, and by total body electrical conductivity as the comparison method in 27 outpatient HIV‐infected men seen at the Houston Veterans Affairs Special Medicine Clinic. Agreement was assessed by comparing the difference between two methods (the bias) with the mean of those two methods. This statistical approach evaluates whether two methods are similar enough that measurements from one might accurately replace those of the other.Results: The mean ± SE for lean‐body mass were 55.98 ± 1.96 kg for total body electrical conductivity and 55.18 ± 1.27 kg for bioelectrical impedance assessment; they ranged from 55.18 ± 1.27 to 63.71 ± 1.89 kg for the prediction equations. Conclusions: In individual subjects, no alternate method gave estimates of lean‐body mass that were the same as estimates from total body electrical conductivity. One prediction equation (Brozek) gave estimates that might be useful for following changes in fat‐free mass over time because the bias did not change substantially for increasing values of lean‐body mass. On the other hand, because there were no statistically significant differences between the mean lean‐body mass estimates by total body electrical conductivity and those measured by bioelectrical impedance assessment or a prediction equation on the basis of body mass index, the latter two methods might be useful in assessing lean‐body mass in groups. (Journal of Parenteral and Enteral Nutrition19:28–32, 1995)