The present invention provides for selection of livestock animals, including bovines, whose genotypes based in the STAT6 gene are correlated with phenotypes reflecting desirable carcass and feedlot traits. These phenotypes include back fat (BFAT), calculated yield grade (CALCYG), cutability (CUT), hot carcass weight (HCW), dry matter intake (DMI), days on feed (DOF), back fat rate (BFAT RATE) and average daily gain (ADG), based on the knowledge of the STAT6 genotypes. The predictive value is based in part on the discovery that certain single nucleotide polymorphisms (SNPs) within the STAT6 gene are linked to phenotypes of economically these important carcass and feedlot traits. Also provided are SNPs within the STAT6 gene useful in reliably distinguishing between a Bos taurus and a Bos indicus bovine. The invention provides methods and compositions for determining STAT6 genotypes and for screening livestock animals to predict which animals will have desirable carcass traits and feedlot traits, allowing producers to selectively breed and manage animals based on desired characteristics, thereby maximizing productivity and profitability in commercial meat production operations.