A limited reproducibility has been ascribed to 24-h ambulatory blood pressure monitoring, especially in relation to the dipper and nondipper phenomena. This study examined the reproducibility of 24-h ambulatory blood pressure monitoring in three recordings of pressure at intervals of 8-15 days in 101 study participants (73% treated hypertensive patients) residing in the city of Recife, Pernambuco, Brazil. SpaceLabs 90207 monitors were used, and the minimum number of valid measurements was 80. No significant differences were found between the mean systolic and diastolic pressures, between the second and third recordings when the normotensive and hypertensive patients were assessed jointly (P=0.44). Likewise, no significant differences were present when the normotensive patients were analyzed separately (P=0.96). In the hypertensive group, a significant difference existed between only the first and second ambulatory blood pressure readings (135.1 vs. 132.9 mmHg, respectively; P=0.0005). Regarding declines in pressure during sleep, no significant differences occurred when continuous percentage values were considered (P=0.27). The values obtained from 24-h ambulatory blood pressure monitoring are reproducible when tested at intervals of 8-15 days. Small differences, when significantly present, always involved the first ambulatory blood pressure monitoring. The reproducibility of the dipper and nondipper patterns is of greater complexity because it considers cutoff points rather than continuous ones to characterize these states.