With aging, arterial stiffness increases and results in cardiovascular diseases. Recently, high brachial-ankle pulse wave velocity (baPWV), measured using a new noninvasive device to estimate arterial stiffness, was reported to be associated with the prevalence of cardiovascular diseases. The purpose of this study was to clarify the association between baPWV with 3-year mortality in community-dwelling older adults and to determine the cutoff value of baPWV in terms of mortality. A total of 530 subjects aged 65 years or older (men/women, 207:323; mean age, 76 years) participated. They were dichotomized by the median value of baPWV. Within 3 years, 30 deaths occurred, including 11 cardiovascular deaths. The high-baPWV group had a higher incidence of total deaths (high-baPWV group vs. low-baPWV group, 8.3 vs. 3.0%, respectively) and cardiovascular deaths (high-baPWV group vs. low-baPWV group, 3.8 vs. 0.4%, respectively). A high-baPWV level was associated with an increased risk of 3-year total mortality after adjustment for age, sex and systolic blood pressure (hazard ratio for high baPWV vs. low baPWV=2.98, 95% CI=1.25-7.07) and with an increased risk of 3-year cardiovascular mortality (hazard ratio for high baPWV vs. low baPWV=10.01, 95% CI=1.21-82.49). A receiver-operating characteristic curve showed that the optimal cutoff value of baPWV for total mortality was 19.63 m s(-1), and for cardiovascular mortality it was 19.63 m s(-1). This study provides a preliminary finding that assessment of arterial stiffness by baPWV might be a useful method to predict mortality risk in community-dwelling older adults. Large longitudinal studies for extended periods of time are necessary to confirm the association.