Research in biometrics suggests that the time period a specific trait is monitored over (i.e. observing speech or handwriting “long enough”) is useful for identification. Focusing on this aspect, this paper presents a data mining analysis of the effect of observation time period on user identification based on online user behavior. We show that online identification accuracies improve with pooling user data over sessions and present results that quantify the number of sessions needed to identify users at desired accuracy thresholds. We discuss potential applications of this for verification of online user identity, particularly as part of multi-factor authentication methods.