Eighty-one untreated elderly patients with clinic-defined isolated systolic hypertension (ISH) and 39 normotensive elderly subjects underwent 24-h ambulatory blood pressure monitoring. Before the ambulatory blood pressure monitoring, EDTA-anticoagulated venous blood was obtained from seated subjects for determination of plasma renin activity. Ambulatory blood pressure and heart rates were determined at 15-30-min intervals by a validated, portable non-invasive technique (Spacelabs 5200). Ambulatory blood pressure variability was defined for each subject as the standard deviation and the coefficient of variation of the ambulatory blood pressure. The mean awake systolic blood pressure was much lower than the clinic-determined value in the ISH group (P less than 0.001), but only slightly so in the normotensive group. Forty-two per cent of the clinic-defined ISH group had mean awake ambulatory systolic blood pressures below the 90th percentile of the normotensive group. A discrepancy between office and ambulatory blood pressures was not associated with blood pressure variability, heart rate or plasma renin activity.