A total of 1,399 persons, between 45 and 79 years of age, who had been identified in a clinic-based case-control study were invited to participate in a follow-up study. The follow-up study was designed to estimate the incidence and progression of cortical, nuclear, and posterior subcapsular cataracts and to evaluate the usefulness of the Lens Opacities Classification System II in a longitudinal study. Survival analyses were performed on 1,193 persons with at least three visits, by using data obtained from Zeiss slit-lamp and Neitz retroillumination lens photographs. The three-year cumulative incidence for persons age 65 to 74 years (the largest age group) was 18%, 6%, and 6% for cortical, nuclear, and posterior subcapsular cataracts, respectively. Progression was much higher than incidence for each type of opacity. Regression, which probably comes mostly from misclassification in the gradings, was modest for cortical and nuclear cataracts but was sizeable for posterior subcapsular cataracts. Patient age, baseline lens status, cataract grading system, definition of change, and analytic methodology can have important effects on estimates of cataract incidence and progression.