In the past few years non-invasive blood pressure monitoring has become increasingly popular in the belief that this approach: can provide accurate mean blood pressure values over a 24-h period or through the day and these values may define better than the casual values the severity of hypertension and its related risk of developing cardiovascular morbidity and mortality. In this paper a number of problems concerning non-invasive ambulatory blood pressure monitoring in hypertension will be reviewed. First, evidence will be shown that blood pressure measurements at intervals up to 30 min can provide a 24-h blood pressure mean similar to that obtained by continuous analysis of the blood pressure tracing, which demonstrates that intermittent readings do not represent a limitation of automatic blood pressure monitoring. Then, it will be shown that the periodical cuff inflations of automatic or semi-automatic blood pressure monitoring may not trigger an alarm reaction or disturb patients' sleep. However, these advantages must be balanced against other still unverified aspects of these new techniques: the inability of intermittent blood pressure readings to evaluate accurately blood pressure variability, which may be a determinant of the overall risk profile, the error inherent in non-invasive measurements of blood pressure and the limited prospective evidence that average 24-h or daytime blood pressure values indeed correlate with the development of target organ damage better than casual blood pressure values do.