With cyber attacks and computer hacking becoming a growing concern in the United States and abroad, measures that could reduce the frequency of these crimes and mitigate their consequences are of a growing interest. To that end, criminological research should be dedicated to understanding the etiology of cybercrime and finding ways to prevent it and to alter the behaviors of its perpetrators. This thesis adds to this body of literature by examining how computer system trespassers may alter their online behaviors when their presence has been detected on the system. Using an experimental design to examine how detection cues presented as a series of messages in an attacked computer system influence the trespassers decisions’ to enter specific commands to influence or identify aspects of the system, it was hypothesized that there would be differences in the commands entered to attack and explore the system, and in IP Addresses used between treatment and control groups. The results do not support most of the hypotheses, but several implications are still drawn from this study.