Ecological vulnerability assessment serves as the foundation for regional development planning and ecological governance and restoration. Jingyuan County in the southern water source area of Ningxia is a national key ecological function zone and also the water supply area for the water diversion project in the central and southern part of Ningxia. Ecological security is of vital importance here. Conducting an ecological vulnerability assessment for this area, revealing its spatio-temporal evolution characteristics, and providing theoretical support for regional ecological environment protection and construction as well as the rational development and utilization of ecological resources are of great significance. This study takes small-scale townships and grids as research units, and based on the ecological exposure-ecological sensitivity-ecological resilience (VSD) model, selects 16 indicator factors in combination with typical ecosystems in the water source area to construct an ecological vulnerability assessment index system. The Analytic Hierarchy Process (AHP) combined with the entropy weight method is employed to evaluate the ecological vulnerability by using the standardized and weighted data through the ecological vulnerability comprehensive index. The variation coefficient, global Moran's I index, and LISA cluster map are utilized to quantitatively analyze and discuss the distribution pattern of ecological vulnerability changes in the Jingyuan water source area.The main research conclusions are as follows: The ecological vulnerability of the study area shows a distribution feature of "low in the east and west, high in the middle"; the average ecological vulnerability index of the four periods shows a decreasing trend, indicating a steady improvement in the ecological environment quality; the ecological vulnerability shows significant positive spatial correlation and spatial aggregation, with low-low aggregation and high-high aggregation being the most obvious; during the study period, the ecological vulnerability index of each township decreased, and the transformation from moderate and severe vulnerability to mild vulnerability was the main trend.