The late frost injury of winter wheat usually occurs during the jointing-heading stage and may result in severe yield loss in large areas, thus it is of significant importance to monitor and assess late frost injury of winter wheat real-timely and accurately. The simultaneous heat and water (SHAW) model is a detailed process model of heat and water movement in plant—snow—residue—soil system, and it has the capability to simulate heat and water transfer within the canopy. The SHAW model was applied to simulate air temperature within winter wheat young ear layer at the sensitive period after jointing stage in Shangqiu City on the basis of field experiment in 2015, and it was also adopted to monitor occurrence and damage level of late frost injury combining with the days after jointing stage. The results indicated that the air temperature within young ear layer (20~60cm) was accurately simulated as a whole, in which about 44.7% and 72.5% of the absolute errors of simulated value were less than 1℃ and 2℃, respectively, and the simulated air temperature at night was better than that in the daytime. Compared with the minimum air temperature measured at the height of 1.5m at the meteorological station, the simulated minimum air temperature within the young ear layer of winter wheat and the low temperature duration can well express the low temperature environment of young ear when late frost injury was occurred. There were large differences among the minimum air temperature measured at the meteorological station, the minimum air temperature measured at 2m height in the winter wheat fields and within the young ear layer because of the influence of field microclimate. The method which transformed air temperature data from the meteorological station into air temperature data at 2m height in the winter wheat fields was better than the method which used air temperature data from the meteorological station as driving data of SHAW model directly, the minimum air temperature simulated by the former method was close to the measured one, and the late frost injury level evaluated by using the former method was in good agreement with the field surveyed one. Therefore, using SHAW model to monitor late frost injury of winter wheat is feasible and applicable, and compared with the traditional monitoring index of air temperature data from the meteorological station it can enhance the accuracy for monitoring the occurrence and damage level of late frost injury.