We studied the relationship between age and prevalence, duration and incidence of clinical and laboratory evidence of ocular Chlamydia trachomatis infection in a cohort of Gambian subjects examined bi-weekly for 6 months. The duration of disease and infection, estimated by stratified survival analysis, proportional hazards regression and Weibull modelling, was markedly age-dependent. The estimated median duration of disease was 13·2 weeks in 0–4-year-old subjects and 1·7 weeks in those age 15 and over. Adjustment for multiple infections, and for missing observations did not alter this trend. The cumulative incidence rate of disease was reduced threefold with age. More rapid disease resolution is the main source of reduction in prevalence of active trachoma and ocular C. trachomatis infection with age; disease incidence was reduced to a lesser extent. This age-dependent resolution may be effected by adaptive cellular immune mechanisms. Mechanisms responsible for natural immunity should receive appropriate emphasis in vaccine design.