Twenty-two animals, eight from genetic improved Nellore breed, six non-improved Nellore and eight from genetic improved Caracu breed, were used to evaluate and to validate the VIÇOSA, CNCPS (level 1 and 2) and NRC (level 1 and 2) systems, for diet formulations. The animals were confined with average live weight of 404 kg to genetic improved Nellore, 345 kg to non-improved Nellore and 434 kg to genetic improved Caracu breed, all with 18 months of age. The forage used was corn silage in forage to concentrate ratio of 50:50 in the diet. The slaughter criterion was determined by ultra-sound and it was executed when the animals reached an average of four millimeters of subcutaneous fat. The Student t test was used to compare the dry matter intake (DMI) and average daily gain (ADG) observed in the genetic groups and those predicted by the VIÇOSA, CNCPS and NRC systems. The VIÇOSA system showed a good estimate for ADG of the animals of genetic improved Nellore and non-improved Nellore breed, but the values were different of observed to improved Caracu. The VIÇOSA system does not have a good estimate for DMI for none of the three genetic groups. The CNCPS system, level 1 and 2 was efficient to estimate the DMI of the three genetic groups; however the estimative for ADG differed from the observed for both level 1 and level 2 of the system and the values were underestimated. In the NRC system, only the level 2 showed predicted values similar to those observed for both DMI and ADG.
diets formulation; intake; selected Nellore; weight gain
