Purpose: To investigate the epidemiology of Gunn's dots and their associations in a population-based cohort of children. Methods: Red-free fundus photographs from 2,286 children aged 11 years to 14 years from the Sydney Myopia Study were graded. Gunn's dots were manually marked and counted within a 6 mm grid centered on the optic disc. Results: One or more Gunn's dots were seen in at least one eye in 82.6% of children. The median number of Gunn's dots per eye was 46 (range 0-482). Most Gunn's dots were found inferior and superior of the optic disc (49.3% and 45.8%, respectively, of the total number of Gunn's dots in the population). The odds for having 1 or more Gunn's dots were 3-fold greater in children with dark brown irides compared with children with blue irides (odds ratio 2.99, 95% CI 1.81 to 4.94, P, 0.0001 adjusted for age, sex, retinal nerve fiber layer thickness, refraction, ethnicity, and axial length). In the same analysis, the presence of 1 or more Gunn's dots was less frequent in children with thin retinal nerve fiber layers (first quartile) compared with children with thick retinal nerve fiber layers (fourth quartile) (odds ratio 0.68, 95% CI 0.49-0.93, P = 0.016). No disease or disease-marker associations were identified. Conclusion: Gunn's dots were visible on fundus photographs in most of the school children, most of the dots being located inferiorly and superiorly to the optic disc. Fundus photographically visible Gunn's dots were associated with darker irides and thicker retinal nerve fiber layers. Gunn's dots are a common and apparently harmless finding in children. Assessment of their clinical significance will require long-term follow-up.