This study is designed to compare the predictability of lean body mass, as measured by whole-body 40K spectrometry, from skinfolds, circumferences and skeletal widths in children 7 to 12 years of age. The specific skinfold sites were back, upper arm, side, waist, abdomen, and calf; the circumference sites were forearm, upper arm (flexed), wrist, thigh, calf, and chest (deflated); skeletal widths included wrist, knee, ankle, elbow, shoulder and hip. In a group of 163 boys, three skinfolds and body weight accounted for 89.7% of the variation in LBM, two circumferences and height and weight accounted for 87.2% of the variation in LBM, and two skeletal widths and height and weight accounted for 87.4% of the variation in LBM. Combining all measurement variables into one analysis resulted in five significant variables: weight, side skinfold, abdomen skinfold, forearm circumference and chest circumference with the coefficient of determination 90.6%, only slightly higher than with weight and three skinfolds. The significant variables from the combined analysis were then used to predict LBM in five separate age groups of boys and a sample of 44 girls. In general, weight, forearm and chest circumference contributed positively to LBM and side and abdomen skinfolds contributed negatively. The regression coefficients for each site were not significantly different among age groups. LBM in children can be estimated from skinfolds, circumferences or skeletal widths with considerable success, as has been shown to be the case in college-age adults.