Numerous trials provide support for the Body Project, an eating disorder prevention program wherein young women with body image concerns critique the thin ideal. Despite medium to large effects, some participants subsequently develop an eating disorder, suggesting that intervention or recruitment procedures could be improved. This study investigated baseline and acute intervention predictors of DSM-5 eating disorder development during a 3-year follow-up among Body Project participants. Combined data from two trials compare participants who experienced eating disorder onset during follow-up (n = 20) to those who did not (n = 216). Participants who did versus did not develop an eating disorder started the intervention with higher eating disorder symptoms (η (2) = 0.08), negative affect (η (2) = 0.06), thin-ideal internalization (η (2) = 0.02), and body dissatisfaction (η (2) = 0.02); the same baseline predictors of eating disorder onset emerged in controls. Attenuated pre-post reductions in eating disorder symptoms (η (2) = 0.01) predicted eating disorder onset but not after controlling for baseline levels. Given that Body Project and control participants who later developed an eating disorder started with initial elevations in risk factors and eating disorder symptoms, it might be useful to develop a more intensive variant of this program for those exhibiting greater risk at baseline and to deliver the prevention program earlier to prevent initial escalation of risk. The fact that nonresponders also showed greater negative affect and eating disorder symptoms suggests that it might be useful to add activities to improve affect and increase dissonance about disordered eating.