Traditional sphygmomanometric blood pressure measurements may lead to errors in the diagnosis of arterial hypertension due to a number of factors, among which are the alerting reaction and pressor response induced in patients by the doctor's visit. This phenomenon, as quantified in our laboratory by continuous intraarterial recordings, is responsible for an average rise in systolic and diastolic blood pressure of 27 15 mm Hg , a rise that does not seem to be reduced by simply desensitizing the patient by means of more frequent physician visits. Twenty-four hour ambulatory blood pressure monitoring may theoretically improve the diagnostic approach to hypertensive patients by overcoming some limitations of isolated cuff measurements. In recording intraarterial blood pressure in 108 ambulant hypertensive subjects, we have found that 24-hour blood pressure values are able to discriminate among patients with different degrees of target organ damage better than isolated sphygmomanometric readings. Moreover, these studies have indicated that 24-hour blood pressure variability may be as important as blood pressure mean values in the assessment of cardiovascular complications. In clinical practice, however, intraarterial blood pressure monitoring is not feasible, and only noninvasive recorders can be used. Use of these devices does not induce any alerting reaction in the patients and does not interfere with day-night blood pressure changes. Although it is characterized by intermittent readings, this approach is not incompatible with a precise estimate of 24-hour blood pressure mean values. However, the discontinuous measurements provided by the presently available noninvasive devices do not allow the precise assessment of blood pressure variability and these measurements are characterized by a considerable degree of inaccuracy when used in ambulant subjects. These limitations, combined with the lack of prospective demonstrations of the diagnostic superiority of ambulatory blood pressure monitoring over office blood pressure readings, still suggest caution in recommending a generalized use of this nevertheless promising approach to the clinical management of hypertension.