In the era of digital libraries, WWW-based electronic databases have become important resources for education and research. Investigating into and then comprehending the behaviors of users in using these systems is very valuable. Based upon the findings of user behaviors, system developers can improve system features, such as query capability and user interface. In conducting user training, librarians can easily locate those system features unfamiliar to users. This paper analyzes and explicates the web server logs of the Taiwan Elsevier SDOS E-journals system, one of the largest and most heavily used full text STM databases worldwide. Results include characteristics of user sessions, server loading, query complexity, query terms, query refinement, browsing, and full-text downloading.