Prevalence data for primary open angle glaucoma (POAG), taken from eight population surveys, was smoothed by curve-fitting to derive composite estimates with respect to quinqennial age groups from 40-44 to 85-89 years. These were applied to national population figures to provide a distribution of cases with respect to age. Estimated prevalence for age 40-89 years in mainly white Caucasian people was 1.2%, rising from 0.2% for those in their 40s to 4.3% for those in their 80s. Of the total cases, 7% were less than 55 years old, 44% were aged 55-74 years, and 49% were older. 'Implied incidence' was estimated from the prevalence results, being 0.11% per year in people aged 55 to 74 years. The analysis applied to relatively narrow definitions of POAG. If 'probable' cases and also 'ocular hypertensives requiring treatment' (of relevance for glaucoma screening) were included, the prevalence would be almost twice as high. Also, a larger proportion of potential cases for a screen would be less than 55 years old, partly because the average age of incident (i.e., newly developed) cases is less than that of prevalent (i.e., all existing) cases.