The major objective of this research was to develop equations to estimate BW and body composition using measurements taken with inexpensive instruments. We used five groups of chickens that were created with different genetic stocks and feeding programs. Four of the five groups were from broiler genetic stock, and one was from sex-linked heavy layers. The goal was to sample six males from each group when the group weight was 1.20, 1.75, and 2.30 kg. Each male was weighed and measured for back length, pelvis width, circumference, breast width, keel length, and abdominal skinfold thickness. A cloth tape measure, calipers, and skinfold calipers were used for measurement. Chickens were scanned for total body electrical conductivity (TOBEC) before being euthanized and frozen. Six females were selected at weights similar to those for males and were measured in the same way. Each whole chicken was ground, and a portion of ground material of each was used to measure water, fat, ash, and energy content. Multiple linear regression was used to estimate BW from body measurements. The best single measurement was pelvis width, with an R2 = 0.67. Inclusion of three body measurements in an equation resulted in R2 = 0.78 and the following equation: BW (g) = -930.0 + 68.5 (breast, cm) + 48.5 (circumference, cm) + 62.8 (pelvis, cm). The best single measurement to estimate body fat was abdominal skinfold thickness, expressed as a natural logarithm. Inclusion of weight and skinfold thickness resulted in R2 = 0.63 for body fat according to the following equation: fat (%) = 24.83 + 6.75 (skinfold, ln cm) - 3.87 (wt, kg). Inclusion of the result of TOBEC and the effect of sex improved the R2 to 0.78 for body fat. Regression analysis was used to develop additional equations, based on fat, to estimate water and energy contents of the body. The body water content (%) = 72.1 - 0.60 (body fat, %), and body energy (kcal/g) = 1.097 + 0.080 (body fat, %). The results of the present study indicated that the composition of a chicken's body could be estimated from the models that were developed.