The study of the genomic variation in cattle has been undertaken to identify loci that allow the prediction of cattle susceptible to Mycobacterium avium ssp. paratuberculosis (MAP) infection. The application of this information is to assist in disease prevention through selective breeding of cattle that are less likely to become infected should MAP exposure occur. A genetic predisposition to MAP infection in cattle was established through differences in prevalence of MAP infection found across breeds and across sires. Heritability estimates for susceptibility to MAP infection have been computed from family and genomic studies and typically range from 0.01 to 0.18, although higher estimates have been reported. Candidate gene and genome wide association studies have both been used to identify the loci associated with MAP infection. Phenotypes of susceptible and non-susceptible cattle have been based on serum ELISA, milk ELISA, fecal culture, tissue culture, PCR of tissue or feces or a combination of these diagnostic methods. The disparity in the sensitivity of the diagnostic methods affects the reliability of the controls, and the stage of MAP infection at which the animal is diagnosed. These differences make comparisons across studies difficult, as genomic studies are sensitive to phenotypic differences. Consequently, few studies have identified the same susceptibility loci in independent cattle populations. This lack of agreement across studies is most likely due to small sample sizes (most studies have less than 500 cases and 500 controls), the lack of consistency in the definition of phenotypes, and the polygenic nature of the disease which could result in different genes having a different level of effect within and across breeds. Identifying loci associated with MAP infection offers a permanent preventative strategy to lower the prevalence of bovine paratuberculosis and to have a better understanding of the disease etiology. To obtain concordant results, larger studies with a consistent phenotype are needed and must be followed with independent validation and functional studies. This will allow genomic prediction to deliver its translational promise to assist in preventing MAP infection in cattle.