Abstract: The study on the changing trends and driving factors of the Normalized Difference Vegetation Index (NDVI), an important indicator to reflect vegetation growth, is of great significance for maintaining regional ecological security. Taking Danjiangkou Reservoir area as the research object, this paper constructed a 1 km resolution NDVI series from 1982 to 2022, by using GIMMS NDVI and MODIS NDVI data based on pixel-level linear regression analysis. These data were used further to analyze the temporal and spatial characteristics of NDVI in the study area. The spatial heterogeneity of vegetation and the driving factors were explored using the geographical detector. The results showed that: (1) The annual mean NDVI increased at a rate of 0.002 6 per year. The spatial distribution of NDVI was characterized by higher values in the northwest and lower ones in the southeast. The areas of low-value were concentrated in low-altitude areas such as Hanzhong, Ankang and Shiyan. (2) NDVI within the reservoir area generally showed an improvement, but with a local degradation. The area with a significant improvement accounted for 98.23% of the total area, which was widely distributed in the whole research area. The degraded area accounted for less than 1%. (3) Annual potential evapotranspiration, elevation and geomorphology were the dominant factor influencing the spatial distribution of local vegetation. The interaction between sunshine hours and elevation had the greatest influence on the spatial distribution of NDVI. The interactions between the influencing factors showed either a two-factor enhancement or nonlinear enhancement.