Clinical associations between Crohn's disease in humans and Mycobacterium avium subsp. paratuberculosis (MAP) have been suggested but not confirmed. Cattle could be sources for MAP, but little information on MAP prevalence with beef has been reported. Samples of ileocecal lymph nodes and swabs of hides and carcasses from 343 animals at cull cattle slaughtering facilities and 243 animals at fed cattle slaughtering facilities across the United States were analyzed for the presence of MAP. Amplification of genetic sequences detected MAP DNA predominantly on hides and in lymph nodes of samples taken at both types of processing facilities. More than 34% of the cattle at cull cow slaughtering facilities had ileocecal lymph nodes that tested positive for MAP DNA. From these same cattle, hide prevalence was more than twofold greater than the prevalence in ileocecal lymph nodes, suggesting that cross-contamination could be occurring during transport and lairage. The prevalence of MAP DNA decreased during processing, and less than 11% of the carcasses tested positive after interventions in the cull cow processing facilities. Using standard double-decontamination and culture techniques, less than 1% of the postintervention carcasses tested positive for viable MAP at cull cow facilities. In samples from the facilities processing only fed cattle, MAP prevalence of 1% or less was detected for ileocecal lymph node, hide, and carcass samples, and viable MAP was not detected. Based on this study, fed cattle carcasses are unlikely sources of MAP, and carcasses at cull cow plants have only a slight risk for transmitting viable MAP, due to current interventions.