Objective: To study the relation between intake of the antioxidant¯avonoid quercetin and subsequent incidence of cerebrovascular disease (CVA). Design: A cohort study carried out among 9208 Finnish men and women 15 y or more of age and initially free from cardiovascular disease. During a 28 y follow-up period in 1967 ± 1994, a total of 824 cases with CVA were diagnosed. Methods: Food consumption data were collected using a dietary history interview method covering the total habitual diet during the previous year. Results: Quercetin intake was not associated with CVA incidence. The relative risk of CVA adjusted for age, serum cholesterol, body mass index, smoking, hypertension, diabetes, geographical area, occupation and intake of beta-carotene, vitamin E, vitamin C, ®bre, various fatty acids, and energy between the highest and lowest quartiles of quercetin intake was 0.99 (95% con®dence interval (CI) 0.71 ± 1.38) for men and 0.85 (CI 0.60 ± 1.21) for women. In contrast, apples, the major source of quercetin in the study population, showed a signi®cant inverse association both in men and women, mainly due to an association with thrombotic or embolic stroke. The relative risks of thrombotic stroke after further adjustment for quercetin intake were 0.59 (CI 0.35 ± 0.99; P 0.45) and 0.61 (CI 0.33 ± 1.12: P for trend 0.02) for men and women, respectively. Conclusions: The results suggest that the intake of apples is related to a decreased risk of thrombotic stroke. This association apparently is not due to the presence of the antioxidant¯avonoid quercetin.