Abstract: In order to solve the task allocation in dynamic environment,proposed a level task allocation method is proposed,.Appropriate tasks are allocated to agents by using the global allocation method based on fitness.When ta environment changes,the matching issues among tasks and agents are solved through local adjustment based on ability and preferences,which makes each agent be able to choose desired task to be performed according to local information.So the robustness and multi-agent utility of the allocation algorithm are improved.Simulation results show that the method is feasible and effective to solve the task allocation problem in a dynamic environment.