Ambulatory blood pressure (ABP) has only rarely been employed in population studies because of the difficulty posed by the greater complexity of this technique. The cross-sectional studies that have been published, however, have allowed a number of conclusions to be drawn. One, 24h average blood pressure of populations is significantly but not closely related to office blood pressure, which thus can not predict accurately daily-life values of blood pressure. Two, 24h average blood pressure is usually less than office blood pressure, the discrepancy increasing with the increase in office values and being of magnitude several mmHg at the office blood pressure of 140/90 mmHg (systolic/diastolic) Three, ABP in women is somewhat less than that in men and ABP for both sexes increases less with aging than does office blood pressure. Four, a circadian profile of blood pressure consisting in values that are much lower at night than are those during daytime characterizes both sexes and all ages with the possible exception of individuals aged 75 years and more, in whom the nocturnal hypotension appears to be attenuated. A similar attenuation has been found for blacks in comparison with whites. The upper limit of normality of ABP has not yet been defined conclusively, although 24h average values </=125/80 mmHg are almost invariably regarded as normal. Normality of ABP should be defined, however, by longitudinal population studies in which ambulatory values are related to prognosis. One of these studies has already been published and others will be completed in the near future.