In recent years, video traffic has been increasing due to the huge demands of video streaming services. Peer-to-peer (P2P) video streaming services (P2PTV) have attracted attention as a solution to decrease the server load because P2P communication can distribute the video data to peers. However, the random peer selection mechanisms of P2PTV conduce to the worldwide spread of P2PTV traffic, and thus it is a large issue to analyze the characteristics of P2PTV traffic. Several traffic measurements have been studied and revealed the characteristics of P2PTV traffic. However, these studies did not focus on users' behavior that has influence on traffic and also did not construct a P2PTV traffic model. In this paper, we analyzed the characteristics of P2PTV traffic that occur in watching a content to construct the P2PTV traffic model of each user. Moreover, we discussed how to model the P2PTV traffic. As a result, we can revealed several characteristics of P2PTV traffic such as the stable traffic and the bursty traffic. Furthermore, we extracted some statistical information and stated modeling the P2PTV traffic.