This paper proposes a heuristic approach to group the peer to peer (P2P) users for sharing the video on demand (VoD) programs in an efficient way. The underlined grouping scheme characterizes the demand behavior to group correlated users into VoD sharing group so that the transmission efficiency can be improved when compared with that without grouping scheme. As the user connectivity and playing behavior in P2P environment vary from time to time, the VoD sharing groups shall be dynamically adjusted. In order to maximize the grouping effectiveness, the compensation factor is proposed as a heuristic indicator to judge whether a peer is valuable to be grouped together or not. And the compensation factor is measured in chunk area, which is composed of consecutive chunks, to minimize the processing complexity. Experimental simulations were performed to evaluate the performance of peers requesting desired position of the content by using the proposed scheme. Compared with the ungrouped P2P environment, the simulation results illustrate that the proposed grouping scheme can effectively increase the high content penetration rate of a movie so that the on demand process of a specific movie can be achieved within the same group .