To investigate the optimal nitrogen (N) application amount and depth that can simultaneously improve both yield and nitrogen use efficiency, we employed the APSIM-Wheat model. The model was calibrated and validated using field trial data from 2012 to 2017, reflecting traditional wheat farming practices in the study area, along with supporting information on crop traits and soil properties from relevant literature. The results showed that a calibrated APSIM-Wheat model could accurately simulate the phenology, biomass, and yield of wheat in the study area, as well as the dynamic changes in soil water and nitrogen conditions. The calibrated APSIM to simulate wheat yield under different N application strategies which showed that increasing the depth of soil N application, in the range of 0-30 cm, could improve wheat yield and yield sustainability in normal and wet years, while increasing N application increased yield instability. The optimal N application depth and amount were about 20-23 cm and 120-150 kg ha-1 and can be reported directly with according to the quantitative relationships between N application amount, depth, agronomic efficiency of N (AEN) and wheat yield under different annual precipitation conditions. Under the optimal N application conditions, the potential yield, biomass, and AEN were 16.5 %, 5.3 %, and 31.7 % higher, respectively, than those under farmers' conventional practices. These results demonstrate that optimized nitrogen management can significantly enhance wheat productivity and fertilizer use efficiency, with important implications for sustainable agricultural practices.