Abstract: The web user-session can be defined as a set of several TCP connections generated by a single user while surfing the web during a given time frame. An activity period, i.e. session, is terminated by a long silent period. This activity period is comprised of several TCP connections which may be used to transfer data. However, identification of active and silent period is not trivial. Correct identification of session is the main goal of our study. Traditional method used threshold-based mechanism for the identification of web user-sessions which required a priori definition of the threshold value. This method is very sensitive to the threshold value, which is very difficult to set correctly. By using clustering techniques, web user-sessions can be identified without requiring a priori definition of threshold values. This paper is based on the definition and identification of web user-sessions. The main goal of this paper is to exploit the property of clustering techniques to group TCP connections in order to identify web user sessions and to compare the performance with that of the threshold-based mechanism.