Urban rail transit has become the main mode to ease traffic congestion, and urban rail transit hub is the most important area for passenger flow distribution in urban rail transit system. Study on the method of simulation and dynamic evaluation of urban rail transit hub has an important significance for train scheduling, operation organization and risk prevention under the viewpoint of traffic safety and service level. This paper analyzed the system characteristics of urban rail transit hub, proposed agent-based modeling and simulation design, simulation process based on Anylogic, dynamic evaluation index set and classification criteria. Furthermore, Beijing South Subway Station was selected for case study, simulation models which included hub simulation model, passenger flow simulation model and train simulation model were built, and experiment results showed that simulation models had a good performance, which was accordant with the reality scene, simulation error was 0.027. Meanwhile, dynamic evaluation which included distribution efficiency, dynamic service level and emergency capacity during the operation period was researched, and optimizing suggestion was proposed, which could provide the beneficial reference for traffic managers.