Net primary productivity (NPP) is a key parameter for monitoring vegetation growth and an important indicator for evaluating ecosystem health. The Ebinur Lake Basin is an important ecological security barrier of the new Eurasian Land Bridge. Therefore,studying the temporal and spatial variation characteristics and driving factors of NPP in the Ebinur Lake Basin is crucial for ensuring the health and stability of its ecological environment. In this study,the CASA (Carnegie-Ames-Stanford Approach) model was used to estimate the NPP of the Ebinur Lake Basin from 2001 to 2020 based on meteorological data,Normalized Difference Vegetation Index (NDVI) data and vegetation type data. Hotspot analysis,coefficient of variation,and trend analysis were used to reveal the temporal and spatial distribution patterns,as well as the evolutionary characteristics of NPP. Additionally,partial correlation analysis was combined with the optimal parameters-based geographical detector (OPGD) model to identify the influencing factors of NPP. The results showed that:(1) The NPP in the Ebinur Lake Basin fluctuated and increased from 2001 to 2020,with a multi-year average NPP value of 199.67 gC m^-2 a^-1 and a growth rate of 1.83 gC m^-2 a^-1. (2) The spatial distribution of NPP had obvious spatial agglomeration characteristics,and the spatial differentiation of NPP showed a pattern of being surrounded by high and low in the middle. The NPP increased in 61.79% of regions,with significant increases concentrated in oases along rivers. (3) The NPP was positively correlated with meteorological factors,and had the strongest correlation with precipitation. NPP increased and then decreased with the increase of elevation and slope,and no significant difference was observed in NPP with the variation in aspect. There were obvious differences in the NPP values of different vegetation types and soil types. The change of land use had both positive and negative effects on NPP. (4) Land use type and vegetation type were the main factors affecting the spatial and temporal differentiation of NPP in the Ebinur Lake Basin. The effects of different factors showed bivariate enhancement or nonlinear enhancement,and the most suitable range or type of vegetation growth among the factors were determined based on the risk detector. The results of this study can be used to monitor the vegetation growth and ecosystem health in the basin,and provide a theoretical foundation for decision makers to formulate a sustainable development program for the Ebinur Lake Basin.