A population-based survey of 760 people 65-74 years of age was conducted in the municipality of Tierp, central Sweden. The methods used to identify glaucoma suspects were applanation tonometry, automatic perimetry (Competer 350), and binocular assessment of the optic disc. For a diagnosis of definite open-angle glaucoma, a visual field defect was required. Open-angle glaucoma was found at a prevalence of 5.70% (95% confidence interval: 4.12-7.27), which is higher than those figures reported for comparable surveys carried out on white populations. The occurrence of capsular glaucoma may in part explain this finding. Twenty of the 45 open-angle glaucomas were diagnosed before the survey. In this category, capsular glaucoma accounted for 60%, whereas 84% of cases detected during the survey were chronic simple glaucomas. Eight cases (18%) were classified as normal tension glaucoma, all of whom were diagnosed during the survey. Chronic simple glaucoma was significantly more prevalent in males than in females.