Technical Abstract: Water is the most important limiting factor to the intensive wheat-maize double cropping system in the North China Plain (NCP). High use efficiency of irrigation water needs better understanding of crop responses to soil water across different climatic conditions. In this paper, the DSSAT-CERES model was calibrated and validated for a two-year field experiment with various irrigation treatments in a typical wheat (Triticum Aestivum L.)-maize (Zea Mays L.) cropping system in the NCP and then was used to identify the interactions among successive seasons on crop water use, soil water balance and crop yield using historical long-term weather data. The model generally simulated soil water content, evapotranspiration (ET), crop yield and above-ground biomass well in responses to these irrigation schedules under different climate conditions. The model performed better under the well watered conditions (2002–2003) than dry seasons (2001–2002), with a root mean square error (RMSE) of 0.29 Mg ha–1 for wheat yield and 0.57 Mg ha–1 for maize yield in the two years. Long-term (1961–2000) simulation results showed that seasonal rainfalls in maize season and irrigations in wheat season influenced substantially subsequent crop growth, water use and yields. Irrigations before wheat planting can be omitted in more than 80% of the simulation seasons, and the saved water can be applied in later growth stages to increase water use efficiency, which may also reduce water drainage losses by 20–30%. Based on the long-term simulation results, reasonable irrigation strategies for the intensified double cropping system are presented considering crop water use efficiency for the whole cropping system and potentially mitigating negative environmental impacts.