Climate change poses unprecedented challenges to soil microbial communities through increased drought frequency and elevated temperatures, yet the resilience of these communities varies significantly across different land-use systems. This study investigated the response and recovery of soil microbiomes to simulated drought and warming treatments across five major land-use types over three years. Experimental treatments included ambient conditions, drought stress (-50% precipitation), elevated temperature (+3°C), and combined drought+ warming across forest, grassland, agricultural, urban, and restored sites. High-throughput sequencing of 16S rRNA and ITS genes revealed that microbial community resilience varied dramatically among land-use systems, with forest soils showing the highest resistance (89% community similarity maintained) and agricultural systems showing the lowest (43% similarity maintained) under combined stress. Fungal communities demonstrated greater resistance to climate stress than bacterial communities across all land-uses, with fungal: bacterial ratios increasing by 2.3-fold under drought+ warming conditions. Grassland and restored ecosystems showed superior recovery capacity, returning to 85-92% of baseline community structure within one year of stress removal. Functional gene analysis revealed enhanced stress tolerance mechanisms in resilient communities, with genes for osmolyte production, heat shock proteins, and dormancy increasing by 180-350% under stress. Soil enzyme activities declined by 25-65% during stress but recovered more rapidly in diverse land-uses, with forest and grassland systems showing complete recovery within 6 months. Network analysis identified keystone taxa that maintained community stability, including stress-tolerant bacteria (Actinobacteria, Firmicutes) and drought-resistant fungi (Ascomycota). Economic modeling estimated that reduced microbial resilience could cost $125-280 ha⁻¹ yr⁻¹ through decreased ecosystem services and increased management requirements. Machine learning models predicted community resilience with 84% accuracy based on initial diversity, soil properties, and land-use characteristics. These findings demonstrate that land-use management significantly influences soil microbial resilience to climate change, providing critical insights for developing climate-adaptive ecosystem management strategies.