Exploring the variations of soil microbial community structure and function under different vegetation types is of great scientific significance for understanding soil ecological functions, and helps to reveal the interaction mechanism between vegetation and soil microbial communities. In this study, two typical vegetation types were selected with similar altitude, wide distribution, and less interference from human activities in Shennongjia National Park, including coniferous forest dominated by Pinus armandii and deciduous broad-leaved forest dominated by Quercus aliena var. acuteserrata, and analyzed the effects of different vegetation types on soil microorganisms based on high-throughput sequencing and molecular ecological network analysis techniques.The study results indicated that: (1) There were significant variations in the relative abundance of dominant soil bacterial phyla, including Bacillota, Actinobacteria, and Acidobacteria, among different vegetation types (P<0.05). Similarly, significant variations were observed in the relative abundances of the dominant soil fungal phyla, Ascomycota and Basidiomycota, across vegetation types (P<0.05). Additionally, the soil fungal Shannon index in P. armandii forests being markedly higher than in Q. aliena forests (P<0.05). (2) Functional prediction of the soil microbial communities showed that the relative abundance of soil chemoheterotrophic bacteria was significantly higher in P. armandii forest soils compared to Q. aliena forests (P<0.001), while the relative abundance of soil nitrogen-fixing bacteria was significantly lower (P<0.05). In the fungal communities of P. armandii forests, the relative abundance of plant saprotrophs and wood saprotrophs was significantly higher (P<0.001), whereas the relative abundance of ectomycorrhizal fungi was significantly lower (P<0.001). (3) Molecular ecological network analysis indicated that the soil microbial community in Q. aliena forests had higher network total nodes, total links, average degree, and modularity compared to P. armandii forests. (4) Partial Mantel tests revealed that soil available phosphorus and the Shannon index of plant diversity exerted the strongest influence on soil bacterial communities (P<0.01), whereas the plant Shannon index and species richness had the most significant influence on soil fungal communities (P<0.01). Soil pH had the greatest influence on both bacterial and fungal functional groups (P<0.001). Partial least squares path modeling (PLS-PM) results indicated that soil pH might indirectly influence the plant Shannon index and species richness by affecting soil nutrients such as available phosphorus, total nitrogen, and organic carbon, which in turn influence microbial Shannon index and species richness (P<0.05). In conclusion, this study demonstrates significant differences in soil microbial community structure and function between different vegetation types, primarily driven by soil pH, available phosphorus, and plant diversity. These findings have important implications for understanding the plant-soil-microbe interaction mechanisms in forest ecosystems.