Abstract: Abstract: Background To analyze the spatiotemporal evolution and driving mechanisms of landscape ecological risk and coupling coordination in production-living-ecological spaces in Zhejiang Province.Method Based on 2000–2020 land-use remote sensing data, landscape ecological risk indices, coupling coordination models, and spatial autocorrelation were used to examine spatiotemporal patterns. XGBoost-SHAP quantified nonlinear and interaction effects of driving factors.Results (1) From 2000 to 2020, production space decreased by 7.29%, living space expanded by 103.41%, and ecological space increased by 0.71%, with urban living space encroaching on agriculture. (2) Spatial patterns showed“high in the north, low in the south; high in plains, low in mountains,”forming“high-risk high-coordination in plains, low-risk low-coordination in mountains.”(3) Coupling coordination exhibited strong positive spatial autocorrelation: H-H clusters in northern plains, L-L clusters in southwestern mountains.(4) Elevation (21.5%) and NDVI (10.2%) were key drivers, with significant interaction effects below 200 m.Conclusion Elevation-based differentiated management is recommended: optimize land use and restore ecosystems in plains, strengthen protection and realize ecological value in mountains, promoting coordinated development across the province.