Aiming at the problem of urban perimeter control, an iterative learning perimeter control method for traffic sub-region considering disturbances is proposed.Firstly, a discrete macroscopic traffic model is established by introducing disturbance term into the vehicle balance equation of road network.Secondly, based on the repeatability of macroscopic traffic flow, the iterative learning control strategy of boundary intersections is designed, and the tracking error of the system is proved to converge to a boundary under the action of bounded disturbance.Finally, taking the actual road network as an example, the simulation results show that the iterative learning perimeter control method considering disturbances can effectively restrain the influence of different degrees of disturbance on the road network performance and improve the traffic condition of the road network.