Abstract: Incorporating ecosystem health (EH) assessment into ecological function zoning (EFZ) provides a scientific basis for regional ecosystem management. This study quantitatively assessed EH in Qinghai and Xizang (QX) of China during 2000–2023 using the vigor-organization-resilience-services (VORS) model across multiple spatial scales (raster, town, county, and basin), and examined spatial clustering patterns using spatial autocorrelation. EH indicators were then integrated with ecological sensitivity to delineate ecological function zones using self-organizing feature mapping (SOFM), and path analysis was applied to identify dominant drivers of EH across different zones. Results showed that: 1) EH in QX generally improved over time but exhibited pronounced scale dependence and spatial heterogeneity. A stable large-scale gradient of ‘higher in the east and lower in the north’ was observed across all scales, while temporal variations were more evident at finer scales and became attenuated at coarser scales. 2) EH displayed significant positive spatial autocorrelation dominated by ‘High-High’ and ‘Low-Low’ associations, although clustering strength declined over time, especially at finer spatial resolutions. 3) QX was designated as five ecological function zones, named using a three-part scheme (geographic sector-dominant ecosystem type-primary function): North-Alpine desert-Sand fixation zone (Zone I), West-Alpine desert steppe-Ecological fragile zone (Zone II), Southwest-Alpine steppe and meadow-Water sensitive zone (Zone III), East-Alpine meadow-Water yield zone (Zone IV), and Southeast-Tropical seasonal rainforest-Ecosystem services provisioning zone (Zone V). 4) At the raster scale of the entire study area, temperature and proportion of forestland area were the most influential factors associated with EH, while NDVI and the proportion of grassland area played relatively smaller roles. Across zones, the dominant drivers of EH differed, reflecting clear spatial heterogeneity in ecosystem regulation mechanisms. These findings demonstrate the value of integrating EH assessment into EFZ and provide scientific support for differentiated ecosystem management and conservation strategies in QX.