Abstract: The crop coefficient curve is an important parameter for estimating the change of water consumption in growing season, and it plays an important role in water management, such as the simulation of evapotranspiration, irrigation forecasting and irrigation decision-making. Previous crop coefficient studies mainly concentrate on the average value of growth period, while rarely focus on daily changes. A crop coefficient simulation method is in need to improve spring maze irrigation forecasting business in Hetao area. Based on the field moisture test data and meteorological station historical observations, the crop coefficient of spring maize is calculated with water balance method, and a dynamic simulation equation is established considering the variation during growing season. It's then evaluated using results of the United Nations Food and Agriculture Organization (FAO) piecewise linear method, and daily rolling estimation of crop evapotranspiration is achieved. At the same time, the leaf area correction method is put forward to estimate the soil moisture content under the water stress, which can provide basis for the development of maize irrigation forecast. Results show that the crop coefficient of spring maize can be described by three curves of the development process, and the change trend of crop coefficient has nothing to do with the output level, but the variation range increases with the increase of output. Considering heat index can reduce the influence of geographical factors on crop coefficient, the simulation equation of maize coefficient are established based on relative accumulative temperature as time variable after emergence, and the decision coefficient are all above 0.92. The maximum (1.30-1.48) and average (0.831-0.919) maize coefficients of each site are calculated by simulation, results are basically the same as the 3 typical values and interval values got by FAO segmentation method, and the range of averaged relative error during growth period is 3.4%-7.2%. Through the analysis, it is concluded that K_c and relative leaf area index are better described by exponential function, and the calculation method of the standard leaf area index is proposed, which can calculate the crop coefficient in any production condition. The simulated soil moisture is consistent with the measured value with average relative error of 6.3%, and less than 15% for 95.8% circumstance, indicating good application prospects. As the soil moisture supply below 1 m is not considered yet, the model should be improved in the future to explore the calculation method of the lower layer water supply.