Macrolides are the first-line treatment against bovine respiratory disease (BRD), and are also used to treat infections in humans. The macrolide, tylosin phosphate, is often included in the diet of cattle as a preventative for liver abscesses in many regions of the world outside of Europe. This study investigated the effects of administering macrolides to beef cattle either systemically through a single subcutaneous injection (therapeutic) or continuously in-feed (subtherapeutic), on the prevalence and antimicrobial resistance of Mannheimia haemolytica and Enterococcus spp. isolated from the nasopharynx and faeces, respectively. Nasopharyngeal and faecal samples were collected weekly over 28 days from untreated beef steers and from steers injected once with tilmicosin or tulathromycin or continuously fed tylosin phosphate at dosages recommended by manufacturers. Tilmicosin and tulathromycin were effective in lowering (P < 0.05) the prevalence of M. haemolytica, whereas subtherapeutic tylosin had no effect. M. haemolytica isolated from control- and macrolide-treated animals were susceptible to macrolides as well as to other antibiotics. Major bacteria co-isolated with M. haemolytica from the nasopharynx included Pasteurella multocida, Staphylococcus spp., Acinetobacter spp., Escherichia coli and Bacillus spp. With the exception of M. haemolytica and P. multocida, erythromycin resistance was frequently found in other isolated species. Both methods of macrolide administration increased (P < 0.05) the proportion of erythromycin resistant enterococci within the population, which was comprised almost exclusively of Enterococcus hirae. Injectable macrolides impacted both respiratory and enteric microbes, whereas orally administered macrolides only influenced enteric bacteria.