Abstract: With an aim of optimizing the stock policy of supply chain system, the paper studies the entities-oriented simulation model of a dynamic supply chain system, and analyzes the properties, actions, events, and methods of entity state transition. It studies the evolution simulation method for the stock policy of the supply chain system by adopting a genetic algorithm. Combining the simulation model and genetic algorithm, the evolution simulation program for the stock policy of supply chain system is developed with the Delphi software developing tool. Selecting average profit of each day as the optimization goal, an example was analyzed and computed. The example showed that the result gained from the evolution simulation method increased the profit by 10.2% when compared with the traditional method.