Abstract: This study investigated the effects of different application methods on soil microbial diversity and functional microbial structure in silage maize by soil microbial sequencing technology through a field experiment with a two-factor orthogonal design: compound fertilizer (C₁~C₃: 300, 525, 750 kg/hm²) and urea (N₁~N₃: 150, 300, 450 kg/hm²), including a nofertilizer control (CK). The results showed that in bacterial diversity, the Shannon index of the C₂N₃ treatment was significantly lower than that of CK. Simpson index decreased significantly in C₂N₂ and C₁N₃, while the Chao1 index of C₁N₃ was significantly higher than that of C₃N₁ and C₃N₃. The Chao1 and Shannon indices of C₂N₂ were significantly elevated compared to C₃N₂ (CK), and C₁N₂ in fungi. In community structure differences, PCoA revealed significant divergence between fertilized and CK groups in bacteria. Distinct clustering was observed between C₃N₁ and C₁N₁ in fungi. Proteobacteria, Acidobacteriota, and Actinobacteriota were dominant taxa. Ascomycota and Basidiomycota were predominant. LEfSe analysis showed that Acidobacteriota was enriched in C₃N₃, Proteobacteria in C₃N₁, Cyanobacteria in C₂N₃, and Solirubrobacter in C₁N₁. In fungi, Penicillium was enriched in C₂N₂, and Aspergillus flavus in C₃N₂. Tax4Fun functional analysis revealed that carbohydrate metabolism, metabolism of other amino acids, metabolism of cofactors and vitamins, xenobiotics biodegradation and metabolism, signal transduction and environmental adaptation have higher abundance in fertilization treatment groups. FUNGuild functional analysis revealed that the samples had high abundance of saprophytic and pathotrophic-saprophytic-symbiotic fungal trophic groups. Overall, the C₂N₂ treatment (525 kg/hm² compound fertilizer + 300 kg/hm² urea) was the best fertilization method for silage corn in the red soil area.