Abstract: Under the circumstance of China's civil aviation transportation volume increased year by year, a continuous increase in air traffic flow, so traffic control safety risks faced by rising and practical pressure. As the air traffic flow under increasing security risk is generated from an interaction and coupling process, it has obvious characteristics such as nonlinear, open and dynamic. To realize the dynamic early warning management and intuitionist, which reflects the system relevance of risk variables. This paper uses the dynamics system in the changeable and nonlinear model to build dynamic ATC safety risk early warning model. In order to reflect the dynamic nature of the monitoring indicators and systems association, this article constructs ATC safety risk dynamic early warning models on the basis of the system dynamics. The above model looks for the growth of air traffic flow as the starting point and the control risk to define internal indicator variables as terminal. Based on actual research and interviews, this paper uses the Vensim to simulate and analysis. As simulated examples can predict ATC safety risk warning status values effectively and visualize the changes of related factors in the system, it is reasonable to determine the risk threshold of the key early warning indicators. Results show that the use of SD to construct ATC safety risks warning model, which can meet the model of the dynamic and systemic.