PURPOSE: To determine in a population,based study whether parapapillary atrophy is associated with glaucoma. DESIGN: Population,based cross-sectional study. METHODS: The Beijing Eye Study included 4,439 of 5,324 subjects invited to participate (response rate, 83.4%). Mean age was 56.2 +/- 10.6 years (range, 40 to 101 years). Color optic disk photographs (30 degrees) were examined morphometrically. Parapapillary atrophy was divided into alpha and beta zones. Glaucomatous optic nerve atrophy was defined by a glaucomatous optic nerve head appearance. RESULTS: After excluding highly myopic eyes, data from 4,003 (90.2%) subjects entered the statistical analysis. Glaucomatous optic nerve damage was detected in 93 (2.3%) subjects. The beta zone of parapapillary atrophy as a whole and measured separately in four disk sectors was significantly larger and occurred significantly more frequently in the glaucomatous group than in the nonglaucomatous group (beta zone total area, 1.21 +/- 1.92 mm(2) vs 0.32 +/- 0.99 mm(2); P < .001). In multiple regression analysis, area of beta zone was significantly associated with age (P < .001), myopic refractive error (P < .001), and presence of glaucomatous optic nerve damage (P < .001), with no significant difference between chronic open-angle glaucoma (n = 72) and chronic angle,closure glaucoma (n = 21; zone area, 1.20 +/- 0.39 mm(2) vs 1.19 +/- 0.46 mm(2); P = 69). CONCLUSIONS: In a population-based setting, the P zone of parapapillary atrophy is significantly larger and occurs more frequently in glaucomatous eyes than in normal eyes of Chinese adults, with no marked difference between chronic open-angle glaucoma and primary angle-closure glaucoma.