Elevated ankle blood pressure (ABP) may be one of the earliest signs of subclinical atherosclerosis. However, its behavior in different degrees of atherosclerotic vascular damage has not been well characterized. We examined the association of ABP and brachial exercise blood pressure with the incidence of future coronary events. A cohort of 3808 consecutive ambulatory persons (mean age 50 years, 34% women), referred to a symptom-limited exercise test and free of cardiovascular events at baseline, was prospectively followed up for 15 years. Altogether, 383 (80 fatal and 303 non-fatal) incident coronary events occurred. Cox proportional hazards models, adjusting for several conventional risk factors, were used to analyse the independent association of ABP with the risk of an incident coronary heart disease (CHD) event. Persons with normal ankle, brachial resting and brachial exercise blood pressures were taken as the reference group. Other groups were formed on the basis of ankle and exercise blood pressures and compared with the reference group. Even in persons among whom the elevated ABP was the only abnormal finding, the multivariate adjusted hazard ratio (HR) of a future CHD event was significantly elevated (HR=1.60, 95% confidence interval 1.20-2.14, P<0.0001). In general, the HRs were higher for fatal events than for non-fatal events. The measurement of ABP could be an inexpensive and non-invasive tool to detect elevated risk of a CHD event.