The reproducibility of ambulatory, home, and clinic blood pressures was compared in 13 untreated mildly hypertensive and 14 normotensive subjects. Each subject had two sets of daily ambulatory recordings, home self-measured readings (over 6 days), and clinic measurements taken 2 weeks apart. Comparisons over the 2 weeks within and among the methods of measurements were made using a repeated-measures analysis of variance. The results showed that there was no consistent average change in the ambulatory or home pressures and no change in clinic diastolic pressures, but the clinic systolic pressure of the hypertensive subjects dropped 6 mm Hg (p less than 0.05), while that of the normotensive subjects showed no significant change. Test-retest correlations of each of the three methods were similar in magnitude, indicating a similar level of reliability. Test-retest correlations of the ambulatory standard deviations, however, were low, indicating a low reliability of this measure of daily pressure variability. These results suggest that the reproducibility of ambulatory pressures may be as good or better than that of home or clinic measurements. They also suggest that the average ambulatory pressure may be preferable as the measurement in clinical trials, since it may be less influenced by measurement anxiety, particularly in hypertensive subjects.