Cooperative downloading is going to improve the amount of downloaded data to the users of Vehicular Networking (VANET). Pointing at the problems of the low utilization of the spatiotemporal resource in DA and unbalance service of cooperative downloading, this paper proposed a balanced cooperative downloading method, which makes use of Euclidean Distance and Manhattan Distance dynamically as one of the principle to selecting the vehicles according to the number of the clients, aiming to keep balanced downloading. On the premise of ensuing throughput, the system can provide relative balanced service using this method. Simulation results indicate the benefits of the proposed scheme in terms of increasing efficiency and balanced profit.