There is a paucity of literature examining the longitudinal course of insomnia using standardized diagnostic criteria. This study aims to evaluate the persistence, remission, relapse, and incidence rates of insomnia symptoms and insomnia disorders according to the Diagnostic and Statistical Manual of Mental Disorders, 4th and 5th edition (DSM-IV and DSM-5). A total of 398 community dwellers were interviewed annually over two years using the Brief Insomnia Questionnaire, a validated questionnaire for deriving insomnia diagnoses. Normal sleepers were defined according to the DSM-5 quantitative criteria as having insomnia symptoms at most twice per week. Estimates were weighted against population age and sex distribution. Persistence for two consecutive years was 26.3, 26.4, and 23.0% for insomnia symptoms, DSM-IV, and DSM-5 insomnia disorders; remission rate was 55.8, 22.9, and 26.1%, relapse rate was 21.8, 1.3, and 0%, while incidence rate was 62.4, 19.6, and 4.5%. The common trajectories for DSM-IV insomnia disorder were to remain the same (26.4%), followed by a change to insomnia symptoms at the second year (25.7%), and at the third year (17.3%). For DSM-5 insomnia disorder, a change to insomnia symptoms at the second year was the commonest (28.3%), followed by continuing the same (23.0%), and a change to insomnia symptoms at the third year (14.0%). Over a two-year naturalistic follow-up, persistence of insomnia disorder was roughly 25%. Changes from insomnia disorder to insomnia symptoms were common; however, remission only occurred in about 25%, highlighting the chronic course of insomnia, perhaps due to a lack of treatment, under-treatment, or resistance to treatment. Incidence of insomnia symptoms was 62.4%, suggesting a high risk of developing insomnia in the general population.