Liver abscesses in feedlot cattle are detrimental to animal performance and economic return. Tylosin, a macrolide antibiotic, is used to reduce prevalence of liver abscesses, though there is variable efficacy among different groups of cattle. There is an increased importance in better understanding the etiology and pathogenesis of this condition because of growing concern over antibiotic resistance and increased scrutiny regarding use of antibiotics in food animal production. The objective of this study was to compare the microbiomes and antimicrobial resistance genes (resistomes) of feces of feedlot cattle administered or not administered tylosin and in their pen soil in 3 geographical regions with differing liver abscess prevalences. Cattle (total of 2,256) from 3 geographical regions were selected for inclusion based on dietary supplementation with tylosin (yes/no). Feces and pen soil samples were collected before harvest, and liver abscesses were identified at harvest. Shotgun and 16S rRNA amplicon sequencing were used to evaluate the soil and feces. Microbiome and resistome composition of feces (as compared by UniFrac distances and Euclidian distances, respectively) did not differ (P > 0.05) among tylosin or no tylosin-administered cattle. However, feedlot location was associated with differences (P ≤ 0.05) of resistomes and microbiomes. Using LASSO, a statistical model identified both fecal and soil microbial communities as predictive of liver abscess prevalence in pens. This model explained 75% of the variation in liver abscess prevalence, though a larger sample size would be needed to increase robustness of the model. These data suggest that tylosin exposure does not have a large impact on cattle resistomes or microbiomes, but instead, location of cattle production may be a stronger driver of both the resistome and microbiome composition of feces.