Personalized normative feedback (PNF) has been used extensively to reduce alcohol consumption, particularly among heavy drinkers. However, the majority of PNF studies have used only descriptive norms (real or perceived pervasiveness of a given behavior). The purpose of the current study was to explore the efficacy of PNF both with and without an injunctive message indicating approval or disapproval based on the participants' standing relative to other students' drinking levels. This randomized trial evaluated two brief web-based alcohol intervention conditions (descriptive-norms-feedback-only condition versus a descriptive-plus-injunctive-message condition relative to an assessment-only control condition). Participants included 176 students who had reported at least one heavy drinking episode in the past month. Participants completed baseline and follow-up assessments of perceived norms and drinking. Follow-up assessments were completed at 2 weeks post-intervention by 165 (94%) participants. Analyses were conducted using zero-inflated negative binomial regression models. As expected, the descriptive-norms-only condition was effective in reducing drinking among heavier baseline drinkers at follow-up relative to the control condition. However, contrary to expectations, the descriptive-plus-injunctive-message condition did not predict less drinking at follow-up. This study was unique in using an injunctive message as an adjunct to descriptive-norms feedback within the context of drinking. Findings highlight the need for additional research into the role of defensiveness, which may serve as an impediment to using injunctive norms/messages in interventions for problematic substance use and other potentially stigmatizing behaviors.