Nowadays, network congestion exists as one of the biggest problems in Peer-to-Peer (P2P) multicasting applications. Currently available application level multicasting strategies are optimized by transmission delays between the source and destination, load balance between peers, contribution incentive scheme, etc. However, consumption of the transmission resources in the whole P2P system is not evaluated adequately. In practice, although the resources for transmission are not enough to serve the networked applications, considerable transmission resources are wasted. In this study, influencing factors related to transmission resource consumption are analyzed in detail. Then, a transmission resource occupation model is proposed. Based on the model, the average transmission resource occupation of the peers is used as a metric of efficiency of transmission resource. Accordingly, a multicasting tree construction strategy, called Transmission Resource Saving Tree (TRST) is provided. In TRST, the new peer detects transmission resource occupation related factors and chooses the already joined peer with minimum additional resource occupation. By doing this, resource occupation of the whole system is reduced. Contrast simulation experiments are carried out based on transmission delays and jitters obtained from PlanetLab. The results suggest, if TRST is implemented, the P2P multicasting system consumes less transmission resource than other available strategies under the same condition. The results also show that the performance of TRST is stable although network condition is varying all the time. These characteristics make it an ideal strategy for P2P multicasting system, especially where transmission resource saving is essential.