To describe relationships between age and sub-clinical cardiovascular disease, manifest chronic disease, and physical functioning and limitations among persons aged 65 years and older, with emphasis on the "oldest old," those 85 years and older. Observational population-based study. Four U.S. communities: Forsyth County, North Carolina; Sacramento County, California; Washington County, Maryland; and Pittsburgh, Pennsylvania. 5,201 men and women aged 65 years and older. Demographic data; histories of cardiovascular disease (CVD), chronic lung disease, arthritis, diabetes, and hypertension; measures of subclinical disease including arm and ankle blood pressures, internal carotid wall thickness and stenosis, ejection fraction, left ventricular mass, fractional shortening, and diastolic function, electrocardiographic left ventricular hypertrophy and cardiac injury score, forced expiratory flow and volume; functional status including self-reported physical functioning, hearing and sight limitations and health status, and performance-based measures of function. These variables were examined among men and women in three age groups: 65-74 years, 75-84 years, and 85 + years. Subgroups of participants with and without manifest CVD were also examined. In women, the prevalence of CVD and other chronic conditions increased with age, and the highest rates occurred among those 85 years and older. In men, prevalence rates increased between the two younger groups, but the oldest group had lower than expected rates for coronary heart disease, cerebrovascular disease, hypertension, and chronic lung disease. In contrast, there were strong age-related linear trends in most of the subclinical measures of blood pressure, atherosclerosis and pulmonary function and in virtually all measures of functional status in both gender groups across the age range. There was a particularly marked decline in functional status between the two older age groups. While subclinical disease was greater and functional status was poorer among those with manifest CVD, with few exceptions, age-related trends were not significantly different between the two groups. Lower than expected prevalence rates of CVD among those aged 85 years and older, particularly among men, in this study of community-dwelling elderly may represent selection bias or a real plateauing in disease prevalence with age. However, subclinical disease appears to increase and functional status to decline across the age range in both men and women regardless of the presence of CVD. The apparent increase in subclinical disease with age indicates potential for CVD prevention after age 65.