ABSTRACT This study assessed the use of conformation and fatness scores of the EUROP sheep carcass grading system to predict weight and percentage of salable meat from Brazilian market lambs. Data were collected from in vivo, carcass, and retail production from 252 uncastrated lambs. Evaluated models included single regressions, two multivariate models, and one determined by the stepwise procedure. Conformation was moderately correlated with weight of salable meat. Fatness scores were correlated with rump perimeter, carcass width, and thoracic depth with coefficients of −0.33, −0.32, and −0.23, respectively. Body weight was the best single predictor for weight of salable meat and cold carcass yield for percentage of salable meat. All multivariate models for weight of salable meat prediction were significant. Stepwise regression with body weight, leg perimeter, thoracic depth, rump perimeter, and fatness scores predicted 98% of weight of salable meat variation. For percentage of salable meat prediction, stepwise regression with cold carcass yield, leg perimeter, and conformation score was significant. The EUROP conformation and fatness scores can be used in Brazil for the prediction of lamb meat production.