Objective: To investigate the relationship between a history of hypertension, cigarette smoking and alcohol intake and the lifetime prevalence of stroke in the oldest-old population. Design: A cross-sectional study. Subjects: All of the Japanese centenarians in the Okinawa Prefecture (266 men and 1,378 women). Methods: Okinawa Prefectural Government conducted health surveys among all of the centenarians in Okinawa. The variables used for analysis were sex, history of stroke, age at the first diagnosis of stroke, history of hypertension, cigarette smoking and alcohol intake. We used multiple logistic regression analysis taking the history of stroke as the dependent variable. Results: The lifetime prevalence value for stroke was 11.0% in Japanese centenarians. Hypertension was independently associated with an increased lifetime prevalence of stroke (adjusted odds ratio = 2.97 and 95% confidence interval: 2.16–4.08). There was no material relationship between sex, cigarette smoking, oralcohol intake and the prevalence of stroke. When the lifetime prevalence of stroke was divided according to whether stroke had been diagnosed for the first time at the age of 90 years or less or over the age of 90, a significant positive association between hypertension and stroke was more pronounced in centenarians with a diagnosis of stroke at the age of 90 years or less than in those over the age of 90. Conclusions: The findings suggest that hypertension may increase the likelihood of stroke in Japanese centenarians in Okinawa although the association between hypertension and stroke was more pronounced in those having stroke at 90 years or younger.