The fecal microbiome is emerging as an essential component of the gut microbiota and host metabolism, whereas in cattle, fecal microbiome characterization is still needed. Recent evidence indicates that small RNAs, such as miRNAs, may be isolated from feces and involved in host–microbe interactions. In this study, fecal samples were collected from the rectal ampulla of Nelore bulls phenotypic divergent for residual feed intake (RFI). miRNA sequencing and 16S rRNA gene (V3-V4 region) were performed to reveal the associations between host miRNAs and microbiome composition and their relationships with the feed efficiency phenotype. Among the 162 identified fecal miRNAs, seven were more expressed in the inefficient group: bta-miR-27b, bta-miR-30a, bta-miR-126, bta-miR-143, bta-miR-155, bta-miR-205 and bta-miR-196a. Using metabarcoding sequencing, we identified 5,005 bacterial ASVs, and after filtering, we used 357 ASVs in subsequent analyzes. Weighted gene coexpression network analysis (WGCNA) was used to identify miRNA and microbiome interactions. We observed significant correlations between fecal miRNA expression and microbiota composition. The differentially expressed fecal miRNAs were correlated with some taxa as Prevotella, Anaerorhabdus furcosa, Bifidobacterium, Bacillales, Succinispira mobilis, Peptostreptococcaceae and Coriobacteriaceae, suggesting that they may play a role in the expression of feed efficiency-related miRNAs. Our results provide a new perspective for exploring host-microbiome interactions that affect FE traits. Taken together, these results point to miRNAs and taxa identified here as potential regulators of feed efficiency, which may provide the knowledge needed to develop future strategies to manipulate the microbiome.