The invention relates to a method for predicting structure evolution of austenite dynamic recrystallization in thermal deformation of plate-strip steel, which belongs to the technical field of steel rolling. The method achieves the predictions of the transformed fraction, the grain size, the grain morphology and the flow stress of dynamic recrystallization by establishing a dynamic recrystallization physical metallurgical model and a dynamic recrystallization cellular automation model. The method achieves the reproduction of the structure evolution of a metal forming process by a computer, not only can save the experimental cost, but also can accelerate the development cycle of a new type of steel. The developed method for predicting the structure evolution of the austenite dynamic recrystallization in the thermal deformation of the plate-strip steel achieves the quantitative, precise and visual descriptions of the grain morphology, the volume fraction and the grain size of the recrystallization process, and can obtain important parameters of the flow stress and the like, thus the method is significant in further analyzing the evolution law of microstructures.