Evidence has been produced that average 24-h or day-time blood pressure (BP) values provide a better diagnosis of hypertension and a better prediction of its risk than BP values provided by one or few isolated measurements. This has led to a number of attempts to simplify the time-consuming procedure necessary to obtain this information and to identify short periods during the 24 h when average BP might closely reflect the 24-h or day-time average values. In 40 ambulant subjects BP was recorded intra-arterially for 24 h using the Oxford method. The BP signal was analysed beat-to-beat by a computer to obtain the average 24-h mean BP values. This value was compared with the mean BP average values obtained by beat-to-beat analysis of subperiods of the same recording, ranging from 30 min to 12 h in duration. For each duration the analysis was performed on 48 different subperiods, each shifted by 30 min from the preceding one, so that every time interval within the 24 h was considered. Throughout the 24-h period the average mean BP values obtained by the analysis of the 30-min subperiods showed very marked differences compared to the 24-h mean BP average. The differences between the subperiod and 24-h values showed a progressive reduction as the subperiod duration was increased. Their magnitude, however, was still marked for subperiods of 2 and 4 h, and only showed a clear-cut attenuation when the subperiods encompassed 12 h.(ABSTRACT TRUNCATED AT 250 WORDS)