||Digital Library (DL) has facilitated the distribution of multimedia contents in the networked information environment. Designers of DLs have typically focused on development of sophisticated query algorithms and data structuring models. Without thorough understanding of users' information seeking behavior, designers of DLs are unable to address users' information needs. Existing studies do not specifically investigate the searching behavior of users who look for multimedia contents, therefore, we choose peer-to-peer (P2P) networks, where multimedia contents are extensively distributed, to study users' behavior. In this paper, we propose a methodology to overcome the difficulty in tracing individuals' searching activities, the obstacle brought by anonymity, one of the characteristics of P2P networks. We successfblly identify users' search sessions and study their searching behavior at the level of individuals in P2P networks. Our preliminary results show that-users searching for multimedia contents are likely to stick on the same topic and submit successive queries. The results of this study will be instrumental in providing ways of understanding users' information seeking behavior and incorporating human factors into the design of multimedia DLs.