Body composition of large-bodied turkey lines was compared to that of a randombred control population (RBC2) at 8, 16, and 20 weeks of age. The large-bodied lines included a line (F) selected for 17 generations for increased 16-week body weight, a subline (FL) of F selected for four generations for increased shank width, and a subsample (FM) of F in which a new line was being developed for increased amount of leg muscle. The F line originated from the RBC2 line. Body weights of F and FM lines were similar and slightly larger than that of the FL line. Body weight of the F, FL, and FM lines was much greater than body weight of the RBC2 line.The abdominal fat pad, based on actual weight and weight as a percentage of live weight, was similar in the three large-bodied lines. The values (91 to 98 g) for the large-bodied lines were about 2.5 times the values for the RBC2 (35 g), when based on actual weight, while the abdominal fat pad was about twice as large in the heavy lines (.94 to .95) as the RBC2 line (.49) when based on a percentage of live weight. Females had larger abdominal fat pads than males.All of the heavy lines were similar in the percentages of ash, protein, fat, and dry matter of the carcass. Carcasses of the heavier lines were fatter, contained more dry matter, and less protein and ash than carcasses of the RBC2 line. The body composition of males and females did not differ at 8 weeks of age. However, at 16 and 20 weeks of age, female carcasses contained more fat and dry matter but had less protein and ash than male carcasses.Variation in body weight accounted for from 8 to 15% of the variation of different measures of abdominal fat. Correlation coefficients between live body weight and carcass composition were not statistically significant. Variation in the abdominal fat pad can explain from 36 to 48% of the variation in fat content of the carcasses.