OBJECTIVE: To estimate the incidence of rhegmatogenous retinal detachment (RD) in a geographically defined population and to compare the probability of RD in residents after cataract extraction with the probability of RD in residents who did not have cataract extraction. DESIGN: Rochester Epidemiology Project databases were used to perform a retrospective population-based incidence study of RD diagnosed between 1976 and 1995 with cohort analyses of the influence of risk factors on the occurrence of RD. PARTICIPANTS: The population of Olmsted County, Minnesota, participated. MAIN OUTCOME MEASURE: Incidence rates of RD adjusted to the age and gender distribution of the 1990 U.S. white population were measured. RESULTS: Three hundred eleven incident cases of rhegmatogenous RD were identified. The mean annual age- and gender-adjusted incidence rate of rhegmatogenous RD was 17.9 per 100,000 persons (95% confidence interval [CI], 15.9-19.9). For idiopathic rhegmatogenous RD alone, the mean annual age- and gender-adjusted incidence rate was 12.6 (95% CI, 10.9-14.3) per 100,000 persons. Ten years after phacoemulsification and extracapsular cataract extraction, the estimated cumulative probability of RD was 5.5 (95% CI, 3.4-7.6) times as high as would have been expected in a similar group of county residents not undergoing cataract surgery. CONCLUSIONS: Cataract surgery is associated with a significantly elevated long-term cumulative probability of retinal detachment.