A cohort analysis of lung cancer mortality in the Netherlands from 1936–1974 is presented. Apart from small corrections, age-specific lung cancer mortality rates can be expressed as the product of a cohort (generation) factor and an age factor. The cohort factors increase until the 1930 birth cohort, after which there is a rapid decline. The age-risk factors increase smoothly with age. Age-specific incidence curves of past and future years will follow the curve of the age-risk factors provided that the cohort-risk factors are unchanging. The method can be applied to other chronic diseases.