This article presents an initial randomized controlled trial of an event-specific prevention intervention. Participants included 295 college students (41.69% male, 58.31% female) who intended to consume 2 or more drinks on their 21st birthday. Participants completed a screening/baseline assessment approximately 1 week before they turned 21 and were randomly assigned to receive Web-based personalized feedback or assessment only. Feedback included normative information, protective behaviors, and personalized blood alcohol concentration information. A follow-up assessment was completed approximately 1 week after a student's birthday. Results indicated a significant intervention effect in reducing estimated blood alcohol concentration (d = 0.33). The intervention effect was moderated by 21st-birthday drinking intentions, and the intervention was primarily effective among those who intended to reach higher levels of intoxication. Results provide some support for normative information as a mediator of intervention efficacy. Overall results provide support for Web-based personalized feedback as an intervention approach for specific events associated with extreme drinking. (Contains 4 figures and 3 tables.)