Under climate change, drought intensity is escalating, threatening national food security and sustainable development. Xinjiang faces perennial water scarcity, with droughts significantly impacting agricultural production across extensive areas of the region. This study aims to quantify drought-induced yield-reduction risks for winter wheat and to develop tailored water management strategies by integrating drought indices with crop modelling for the Tarim River Basin from 1981 to 2021. The DSSAT-CERES-Wheat model was coupled with the Standard Precipitation Evapotranspiration Index (SPEI) and Soil Moisture Deficit Index (SMDI) to evaluate yield reductions. Adaptive strategies focused on adjusting sowing dates and optimizing irrigation regimes during critical growth stages. The study area experienced persistent drought in 1984-1987 and alternating wet-dry conditions during 2002-2011. Drought frequency ranged primarily from 4 to 12 events. Duration and severity followed SPEI < SMDI0-10 < SMDI10-40, with higher values in Kashgar and Aksu. Winter wheat yield reductions were more severe in SMDI-identified drought years. To effectively mitigate drought-induced yield losses, this study proposes an adaptive management strategy centered on sowing date optimization and supplemental irrigation (30-60 mm) at the jointing stage. This approach demonstrates significant potential for enhancing regional winter wheat productivity while providing a scientific basis and practical guidance for efficient agricultural water resource utilization in arid regions.