The reproducibility of the whole-day blood pressure, measured by an automated non-invasive device at 7.5 minute intervals, was examined in 6 hypertensive patients in a quiet in-hospital environment. During a repeat study performed two weeks after the baseline monitoring period, the average 14-hour systolic blood pressure had changed significantly in 3 patients and the diastolic blood pressure in 2. If the day was divided into 12 consecutive 2-hour periods, systolic blood pressure averages during corresponding periods of the two studies correlated significantly in 3 of the 6 patients, and diastolic blood pressure averages in 5. After a third 24-hour monitoring period, carried out in 4 of the 6 patients, it was found that the baseline study correlated more strongly with the second repeat study than with the first, and that the correlations between the two latter studies were the strongest of all. Thus, even during relatively non-stimulatory conditions the 24-hour blood pressure is not constant in all patients; there is a tendency, however, towards greater reproducibility of the whole-day blood pressure patterns as patients become more familiar with the monitoring procedure.