This study investigated the effectiveness of two interventions in reducing eating disorder risk factors under naturalistic conditions in sororities. Based on previous research, the campus sororities chose to implement a semi-mandatory, two-session eating disorder prevention program to all new sorority members (N=90) during sorority orientation. To facilitate evaluation, sororities agreed to random assignment of new members to either a cognitive dissonance or media advocacy intervention. Undergraduate peer facilitators ran the groups. Although both interventions had an effect, cognitive dissonance generally was superior at eight-month followup. Results further support the utility of cognitive dissonance in reducing eating disorder risk factors, and suggest that non-doctoral level leaders can deliver the program. Results also indicate that a semi-mandatory format does not reduce effectiveness.