The understanding of how nutrition influences the body composition of growing animals has fascinated researchers for centuries. It involves the expertise of scientists with different areas of knowledge, encompassing the composition of the diet and its nutritive value to the fermentation and digestion of substrates to the absorption and metabolism of nutrients, and finally, to the deposition of fat, protein, and minerals in body tissues. The comparative slaughter technique is the preferred method to assess the body composition of growing and finishing animals. However, the methodological procedures are labor-intensive, expensive, and time-consuming, facilitating the incidence of errors and inconsistencies of the measurements that are collected, including the initial animals body composition. First, retained fat and protein (RP) are used to compute retained energy (RE). Then, RP and RE are used to compute protein and energy requirements for growth. Heat production, calculated from the metabolizable energy (ME) intake for animals at maintenance, is used to compute maintenance requirements. Three areas of concern exist for this approach: 1) the efficiencies of possible mobilization of fat and protein tissues during the feeding period are unaccounted for, especially for the animals fed near the maintenance level of intake; 2) the correlation between observed and predicted RP when using predicted RE is higher than when using observed RE (0.939 vs. 0.679); and 3) the disconnection when predicting partial efficiency of use of ME for growth using the proportion of RE deposited as protein carcass approach versus using the concentration of ME of the diet diet approach. These concerns raised questions about the interdependency between predicted RP and RE and the existence of internal offsetting errors that may prevent overall adequacy in predicting energy and protein requirements of beef cattle.