One hundred two animals were evaluated to develop prediction equations for estimating hind-quarter composition from live animal ultrasound measurements and to compare these equations with those developed from carcass measurements. To develop prediction equations used all-possible regressions procedure for variables reductions and stepwise regression. Final models using live animal variables included live weight (PVUS), Longissimus muscle area (ULMA) and fat thickness (UFAT) for weight (PCTT; R² = 0.78) and percent (RCTT; R² = 0.18) of hind-quarter retail product. The models using variables carcass included hot carcass weight (HCW), carcass Longissimus muscle area (CLMA) and carcass fat thickness (CFAT) and Conformation (CONF) for weight (PCTT; R² = 0.89) and percent (RCTT; R² = 0.67) of hind-quarter retail product. Results indicate that composition predictions equations developed from live animals and ultrasound measurements can be useful to estimate yield hind-quarter retail product.
fatness; cattle; Longissimus muscle area; ultrasound; yield retail product