Oilseed flax (Linum usitatissimum L.), an annual plant in the Linaceae family, is a significant oil crop cultivated in the high-altitude arid regions of China. Until now, the Decision Support System for Agrotechnology Transfer (DSSAT) model has not been used to simulate the growth of flax. This study aims to develop a new flax model for DSSAT, based on field data collected from 2020 to 2023 in the arid regions of Dingxi and Yuzhong, China. We utilized the CSM-CROPGRO-Canola as a template for genetic coefficient adjustment, along with modifications to the water and nitrogen balance modules. This model was used to analyze the response of flax yield to weather factors (maximum temperature, minimum temperature, and precipitation) and nitrogen fertilizer application during dry years, normal years, and wet years over the past 30 years. The results indicate that the flax model demonstrates good predictive accuracy with a low root mean square error (RMSE) and high index of agreement (D-index) in simulating key stages such as the anthesis and maturity phases, as well as in predicting biomass, leaf area index, yield, and soil moisture content. In addition, flax yield exhibited varying degrees of sensitivity to maximum temperature, minimum temperature, precipitation, and nitrogen fertilizer under different precipitation years. Therefore, in dry years, it is recommended to prioritize temperature and moisture management while applying nitrogen fertilizer cautiously, while in normal years, a balance between water and nitrogen use should be maintained to optimize fertilization effects. In wet years, nitrogen utilization should be maximized to enhance yields. The yield of flax under different precipitation conditions follows this trend: wet years > normal years > dry years. This study demonstrated the significant practical value of developing a flax model in the arid regions of China.