Observational studies of cattle production systems usually find that cattle from conventional dairies harbor a higher prevalence of antimicrobial resistant (AMR) enteric bacteria compared to organic dairies or beef-cow operations; given that dairies usually use more antimicrobials, this result is not unexpected. Experimental studies have usually verified that application of antimicrobials leads to at least a transient expansion of AMR bacterial populations in treated cattle. Nevertheless, on dairy farms the majority of antibiotics are used to treat mastitis and yet AMR remains relatively low in mastitis pathogens. Other studies have shown no correlation between antimicrobial use and prevalence of AMR bacteria including documented cases where the prevalence of AMR bacteria is non-responsive to antimicrobial applications or remains relatively high in the absence of antimicrobial use or any other obvious selective pressures. Thus, there are multi-factorial events and pressures that influence AMR bacterial populations in cattle production systems. We introduce a heuristic model that illustrates how repeated antimicrobial selection pressure can increase the probability of genetic linkage between AMR genes and niche- or growth-specific fitness traits. This linkage allows persistence of AMR bacteria at the herd level because subpopulations of AMR bacteria are able to reside long-term within the host animals even in the absence of antimicrobial selection pressure. This model highlights the need for multiple approaches to manage herd health so that the total amount of antimicrobials is limited in a manner that meets animal welfare and public health needs while reducing costs for producers and consumers over the long-term.