Abstract: In order to improve timeliness and accuracy of determining the sowing date of spring soybean, the farmland information collection system was used to obtain the soil moisture and environmental information. According to the principle of synchronization between the soybean sowing process and the appropriate environment and soil moisture, the membership function method was used to function the relationship between the sowing agronomic requirements and the sowing influencing factors, and the knowledge model for determining the reasonable sowing date of spring soybean was established. The model validation uses the meteorological and soil temperature and humidity parameters of two different ecological sites, combined with the decision date of the model and the actual sowing date and seedling emergence situation in various regions in the past two years, to validate the model for determining the reasonable sowing date of spring soybean. The results showed that the output results of model for determining the sowing date of Bei15 in 2021 and Bei14 in 2022 were April 23, April 24, May 7 and May 10, respectively. The monthly error is 3.33% and 9.68% respectively, and the accuracy of the model results is 96.67% and 90.32% respectively. The results of the model design are in good agreement and applicability with the current high-yield spring soybean actual farming system, which can provide a technical reference for agricultural producers to determine the sowing date of spring soybean by means of information technology.