While there exist different approaches to evaluate and refine the design of dynamic traffic management systems, a simulation based approach has distinct advantages. In a computer-based laboratory environment, complicated interactions between the components of the system can be studied in a controlled environment. In this dissertation, a simulation laboratory that integrates traffic control and routing strategies with driver behavior and network performance models is developed. The system is designed to facilitate the evaluation and refinement of traffic management systems under various conditions, taking into account the dynamic and stochastic nature of transportation systems. By obtaining suitable MOE's under a wide range of scenarios, one can improve the design of traffic management strategies. In this chapter, we describe a framework of applying the simulation approach for evaluation and design refinement of dynamic traffic management systems.