Abstract: Non - communicable chronic diseases (NCDs), as a primary global health threat, have caused an increasingly heavy burden all over the world. Compared with traditional statistical model, the agent - based model no longer focuses on fitting general characteristics, but to simulate individual changes under the interaction between individuals and environmental influence on individuals by building individual - based models for simulating the characteristics of a population. By including time - dependent changes in characteristics of human and environment and their interactions, the agent - based model could be used to simulate dynamic changing process of changes in characteristics of human and environment with time. Therefore, the agent - based model can be adopted to simulate complex problems closer to the real situation. The study introduced the structure and characteristics of the agent - based model and reviewed the application and limitations of the model in prediction of NCDs prevalence to provide theoretical and empirical support for the application of the model in NCDs prevention and control.