Low blood pressure has been associated with increased mortality in older people, but it is unclear whether the hypotension is a risk in its own right or just a marker of disease. In this study we investigated the extent to which those in the lowest decile of systolic and diastolic pressures could be predicted by measures of cardiovascular disease and frailty. We studied 782 people 70 years and over drawn from the only group of general practices in a rural township. Hypotension was defined separately for systolic and diastolic blood pressures as a pressure less than the tenth percentile for the sample as a whole. This was a systolic pressure of ≥122 mmHg and a diastolic pressure of ≤68 mmHg. There was a significant relationship between low systolic blood pressure and male sex, history of myocardial infarction by questionnaire and low body mass index (BMI), and between low diastolic pressure and male sex, history of angina and myocardial infarction, use of one or more hypotensive drugs, low BMI, low corrected arm muscle area, low self-maintenance score and the use of two or more home services. There was no association with ECG abnormalities. In the multivariate analysis only 4.2°Oof those with low diastolic pressure and none of those with low systolic pressure could be correctly classified. Hypotension in old age is only partially explained by established cardiovascular disease and frailty.