Using a community adolescent sample, we aimed to (a) empirically derive eating disorder (ED) symptom groups, (b) examine the longitudinal stability of those groups over 10 years, and (c) identify risk factors associated with ED group stability and transition through young adulthood. Young people (N = 2,287) from the Project EAT cohort participated at baseline (1998–1999) and at 10-year follow-up (2008–2009). Participants completed anthropometric measures at baseline and self-report surveys on disordered eating symptoms and risk factors at both time points. Latent transition modeling was used to test the first two aims and multinomial logistic regression was used for the third aim. Three groups emerged and were labeled as: (a) asymptomatic, (b) dieting, (c) disordered eating (e.g., binge eating, compensatory behaviors). Stability of group membership over 10 years was highest for those in the asymptomatic group, while those in the dieting group showed equal likelihood of transitioning to any group. There was a 75% chance that those in the disordered eating group would continue to belong to a symptomatic group 10 years later. We found that these transitions could be predicted by baseline risk factors. For example, adolescents with one standard deviation higher depressive symptoms than their peers had 53% higher odds (OR = 1.53, 95% CI 1.09–2.16) of transitioning from the asymptomatic group to the disordered eating group. Transition among ED groups is relatively common during adolescence and early adulthood. By targeting risk factors such as self-esteem and familial factors in early adolescence, prevention efforts may be improved.