Land-use change is a major driver of biodiversity loss and ecosystem transformation in forest landscapes. Different land-use systems frequently exhibit differences in belowground microbial communities and soil properties. Accordingly, this study compared soil microbial abundance, diversity, community composition, predicted functional potential, and soil fertility between adjacent pine–oak forest and coffee plantation sites in the Sierra Norte de Puebla, Mexico. Soil samples were analyzed using conventional microbiological methods, soil fertility assessments, and metagenomic sequencing on the DNBseq platform. Metagenomic analyses revealed differences in the taxonomic composition of bacteria, archaea, eukaryotes, and viruses between the two land-use systems. Taxonomic richness increased from 393 taxa in the pine–oak forest to 428 taxa in the coffee plantation, whereas forest soils exhibited a more balanced microbial community structure and coffee plantation soils showed a greater representation of microbial groups associated with environmental adaptation and nutrient cycling. Functional annotation based on the COG and KEGG databases indicated that broad predicted functional profiles remained largely conserved despite taxonomic differences, reflecting similar distributions of predicted functional categories between the two land-use systems. Soil fertility analyses showed higher concentrations of ammonium nitrogen (15 vs. 4 mg kg−1) and available phosphorus (2.6 vs. 1.5 mg kg−1) in coffee plantation soils, whereas forest soils contained greater organic matter (8.1% vs. 7.0%). Overall, these findings indicate that the coffee plantation site exhibited a microbial community structure that differed from that of the adjacent pine–oak forest, while both systems exhibited broadly similar predicted functional profiles inferred from gene annotations. These observations are consistent with differences associated with land use under the environmental and management conditions evaluated.