PURPOSE: This study explored the development of phobic disorder as young subjects move into late adolescence. The objectives were to determine: the frequency of clinical, subsyndromal and subthreshold phobia in late adolescence; the incident episode and stability of phobic symptoms as adolescents move from early to late adolescence; and the relationship between demographic characteristics in early adolescence and the presence of clinical, subsyndromal and subthreshold phobia in late adolescents.METHODS: The data were taken from a two-stage epidemiological study to determine the frequency of and risk factors for DSM-III major depressive disorders and suicidal behaviors in a community population of adolescents. Data were also collected for the frequency of phobias. Three levels of phobias were defined.RESULTS: The prevalences of clinical, subsyndromal and subthreshold phobias in late adolescence were 0.40%, 1.34%, and 5.74% respectively. The weighted incidence of clinical phobia was 0.28%, of subsyndromal phobia was 0.19% and of subthreshold phobia was 4.63%. Race, sex, and having a subthreshold phobia at the initial interview were significantly associated with subsyndromal phobia in late adolescence. Diagnosis of clinical phobia at initial interview was significant in all the multivariate models. Having a subthreshold phobia at initial interview was significant for subsyndromal phobia in late adolescence.CONCLUSIONS: Data on late adolescence showed a strong trend towards remission of phobic symptoms and disorder. However, those with clinical phobia at baseline had an increased risk of being diagnosed again at late adolescence. The most frequently reported symptoms were simple phobias, suggesting that simple phobias start early, have a higher prevalence, and are more frequent over time than social or agoraphobia. Further longitudinal studies are needed.