Central Asia (CA) is one of the most representative arid regions in Eurasia. Understanding vegetation sensitivity to different water stress is crucial for developing effective drought mitigation strategies. However, knowledge of how CA vegetation responds to different water stresses remains limited. In this study, we used multiple remote sensing vegetation indices and meteorological reanalysis data to investigate vegetation sensitivity to precipitation (Pre), soil moisture (SM) and vapor pressure deficit (VPD) stresses across CA during 1982-2020. Future vegetation sensitivity and its trends were projected using data from the Coupled Model Intercomparison Project Phase 6 (CMIP6). Results show that vegetation was most sensitive to SM stress, with mean sensitivities to Pre, SM and VPD stresses of 0.40, 0.69 and 0.44, respectively. Compared with other land cover types and climate zones, vegetation in forests, wetlands and relatively cold and wet climate zones exhibited slightly lower sensitivity to SM stress but slightly higher sensitivity to VPD stress; however, these areas still showed the highest average sensitivity to SM stress. Irrigation significantly reduced crop sensitivity to SM stress, with sensitivity decreasing as irrigation water use increased. Trend analysis indicates that vegetation sensitivity to SM stress shows a obvious increasing trend during 1982-2020, while sensitivity trends to Pre and VPD stress are not obvious. Trend-shift analysis revealed that 9.27 %, 18.28 % and 7.10 % of vegetation areas experienced significant shifts in sensitivity trends for Pre, SM and VPD stress, respectively. The dominant shift types were D-I (decreasing to increasing) and I-I (increasing to increasing). Future projections suggest that vegetation sensitivity to SM stress will remain the highest and continue to increase overall. These findings enhance understanding of vegetation responses to different water stresses in CA and provide insights applicable to other arid regions worldwide.