Uncertainties in global hydrological and climate models challenge future estimates of crop water use and sustainability.

Assessing crop water requirements and sustainability under future climates is crucial to concurrently address water scarcity, food security, and ecosystem preservation. Yet, global climate and hydrological models differ substantially in estimating precipitation, potential evapotranspiration, and renewable freshwater, challenging future crop water use and sustainability assessments. Here, we perform a multi-model analysis to quantify how model uncertainty propagates into crop water needs and sustainability estimates. Results show end-of-century global uncertainty spreads of ~ 18% and ~ 51% for green and blue water footprints, respectively, relative to ensemble means. Blue water sustainability exhibits even larger variability, from 250% to 451%, depending on scenario and model choices. These uncertainties further intensify at finer spatial scales. Hydrological model discrepancies dominate blue water assessments, whereas climate model variability primarily affects green water estimates, particularly under high-emission scenarios. Accounting for model uncertainty is essential for reliable crop water estimates and to inform agricultural water management under climate change.