Assessing water scarcity risks under climate change has become an important research topic for sustainable development. Regional water scarcity is driven not only by direct local water deficits but also by indirect effects from upstream supply chains. Despite their significance, existing studies seldom integrate both local water scarcity and indirect water scarcity comprehensively. This study utilizes multi-regional input-output tables (MRIO) to quantify virtual water flows among eight provinces in the Yellow River Basin, elucidating the extent of local (WSI) and indirect water scarcity (IWS) from 2007 to 2017. Leveraging Representative Concentration Pathway (RCP) projections and Shared Socioeconomic Pathway (SSP) scenarios, the research further projects future virtual water flow patterns and associated water scarcity risks in the Yellow River Basin from the 2020s to the 2090s. Findings reveal that downstream provinces (Shandong, Henan, Shanxi) experience more severe water scarcity-both locally and indirectly-than upstream regions (Inner Mongolia, Gansu). Local water scarcity surpasses indirect scarcity, with the agricultural sector predominantly driving IWS, accounting for 76.1% to 91.3%. Additionally, downstream provinces facing severe water scarcity not only exhibit high local water use but also rely on imports from middle and upper regions grappling with water shortages. Under SSP1-RCP2.6 and SSP5-RCP8.5 scenarios, water scarcity risks in the Yellow River Basin are projected to intensify, with the overall WSI potentially reaching 0.59 and IWS attaining severe levels of 0.42 by the 2050s. This study enhances the understanding of water scarcity risks in arid and semi-arid regions, providing valuable insights for policymakers to develop more climate-resilient water-resource management strategies.