Dryland ecosystems are highly exposed to intensified drought under climate change, while their capacity to withstand and recover from multi-source drought remains insufficiently quantified. In this study, we develop an Integrated Drought Vegetation Resistance and Resilience Analysis (IDVRRA) approach that couples multi-source drought indices with satellite observations to evaluate vegetation stability in the arid and semi-arid regions of China from 2000 to 2020. IDVRRA employs a vine copula model to explicitly capture the multivariate probabilistic dependencies between multi-source drought indices (SPI, SRI, SSI) and vegetation indicators (NDVI). It could quantify vegetation resistance and resilience, along with their driving factors. The results indicate that NDVI exhibits a significant upward trend, with a higher mean NDVI (0.22) and faster greening (0.0019/a) in semi-arid regions than in arid regions (0.11, 0.0009/a). Despite this greening, vegetation stability is generally moderate to low, with mean coefficients of variation ranging from 0.098 to 0.100. Vegetation resistance is relatively strong, with more than 60% of semi-arid areas classified as highly resistant, whereas resilience is predominantly moderate (0.8 to 1.2) across more than 90% of the study area. Vine Copula analysis identifies the SPI-SRI-SSI combination as the optimal multi-source drought configuration and reveals agricultural drought (SSI) as the dominant driver of vegetation response. The IDVRRA framework offers a quantitative method to analyze nonlinear and lagged vegetation responses to complex droughts, providing a scientific foundation for evaluating dryland ecosystem resilience and guiding ecological restoration and climate-adaptive management in arid regions.