Crop yields (Y) and virtual water content (VWC) of agricultural production are affected by climate variability and change, and are highly dependent on geographical location, crop type, specific planting and harvesting practice, soil property and moisture, hydro-geologic and climate conditions. This paper assesses and analyzes historical (1985–2009) and future (2040–2064) Y and VWC of three cereal crops (i.e., wheat, barley, and canola) with high spatial resolution in the highly intensive agricultural region of Alberta, Canada, using the Soil and Water Assessment Tool (SWAT). A calibrated and validated SWAT hydrological model is used to supplement agricultural (rainfed and irrigation) models to simulate Y and crop evapotranspiration (ET) at the sub-basin scales. The downscaled climate projections from nine General Climate Models (GCMs) for RCP 2.6 and RCP 8.5 emission scenarios are fed into the calibrated SWAT model. Results from an ensemble average of GCMs show that Y and VWC are projected to change drastically under both RCPs. The trade (export-import) of wheat grain from Alberta to more than a hundred countries around the globe led to the annual saving of ~5 billion m 3 of virtual water (VW) during 1996–2005. Based on the weighted average of VWC for both rainfed and irrigated conditions, future population and consumption, our projections reveal an annual average export potential of ~138 billion m 3 of VW through the flow of these cereal crops in the form of both grain and other processed foods. This amount is expected to outweigh the total historical provincial water yield of 66 billion m 3 and counts for 47% of total historical precipitation and 61% of total historical actual ET. The research outcome highlights the importance of local high-resolution inputs in regional modeling and understanding the local to global water-food trade policy for sustainable agriculture.