The dearth of long-term follow-up studies of community-based samples and differences in methodology in existing studies highlight the need for research designed to examine the stability, comorbidity, and diagnostic thresholds of depression and anxiety in the community.