Purpose: To describe the 5-year incidence of retinal microvascular abnormalities and analyze their associations with arterial hypertension in nondiabetic elderly Chinese subjects in urban and rural Beijing in a follow-up of the original Beijing Eye Study 2001. Design: Population-based study. Participants: The Beijing Eye Study 2006 included 3251 subjects who had participated in the Beijing Eye Study 2001 and returned for reexamination. Methods: By using fundus photographs, we determined the 5-year incidence and change (progression or regression) of retinal microvascular abnormalities, including focal narrowing (FN) of retinal arterioles, arteriovenous nicking (AVN), and retinopathy. Arterial hypertension was graded as controlled treated arterial hypertension, uncontrolled treated arterial hypertension, and untreated arterial hypertension. Main Outcome Measures: Incidence of retinal microvascular abnormalities. Results: After excluding patients with diabetes mellitus or any retinal or optic nerve disease, 2058 subjects (1409 women [68.5%]) were included in the study. The mean age of subjects was 53.8 +/- 9.5 years, and the mean refractive error was -0.24 +/- 2.04 diopters. The 5-year accumulated incidence of FN, AVN, and retinopathy was 4.1%, 1.4%, and 3.3%, respectively. Adjusted for age, gender, and region of habitation, the 5-year incidence of FN, AVN, and retinopathy increased 1.58 times (95% confidence interval [CI], 1.24-2.00; P < 0.001), 1.55 times (95% CI, 1.04-2.32; P = 0.03), and 1.46 times (95% CI, 1.12-1.91; P = 0.005), respectively, for every 10 mmHg increase in mean arterial blood pressure. After adjusting for age, gender, and region of habitation, the rate of regression of FN decreased by 28% with every increase in the grade of the arterial hypertension classification. Correspondingly, the rate of regression of FN was significantly higher in the group with controlled arterial hypertension (44.4%) than in the uncontrolled (22.6%) and untreated (11.5%) groups (P = 0.01). Conclusions: Arterial hypertension was strongly associated with the incidence and longitudinal change of retinal microvascular abnormalities in nondiabetic subjects. Better control of arterial hypertension was correlated with a decreased incidence of retinal microvascular abnormalities and an increased rate of regression of FN. This suggests that retinal FN as potentially the first step of retinal microvascular abnormalities is reversible if arterial hypertension is controlled. Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Ophthalmology 2012;119:2592-2599 (C) 2012 by the American Academy of Ophthalmology.