OBJECTIVE: To examine the extent to which one session of repeated automated (30-45 min) blood pressure measurements in non-treated, hypertensive patients can be used to predict daytime ambulatory blood pressure (ABP). METHODS: Two hundred untreated black hypertensive patients (mean age 50.2+/-11.2) were retrospectively stratified on the basis of the daytime ambulatory diastolic blood pressure (DBP) into mild, moderate and severe hypertensives (90 mmHg <DBP < 140 mmHg). Office blood pressure was determined at 3 min intervals for a total of 30-45 min using the Dinamap Critikon 1846 SX monitor. Twenty-four hour ambulatory blood pressure (ABP) was performed on the same day in all patients using a SpaceLabs 90207 device. RESULTS: The mean diastolic Dinamap measurements compared with the mean ambulatory daytime DBP were very similar in terms of 'traditional' statistics (mean: 114 +/- 14.6 versus 113 +/- 12 mmHg, correlation: 0.80). However, agreement analysis according to Bland and Altman showed a poor agtreement with an error as high as 19 mmHg. The mean systolic Dinamap measurements compared with the mean ambulatory daytime SBP were significantly overestimated (mean: 185 +/- 25.0 versus 170 +/- 19.3 mmHg, correlation 0.82) and the error was as high as 43 mmHg. On residual analysis, 75% of the diastolic Dinamap values and only 35% of the systolic Dinamap measurements were within an error of +/- 10 mmHg. The average of the first five Dinamap measurements had the greatest predictive ability for the daytime ABP. CONCLUSION: As a result of the poor agreement between the two methods, we suggest that, from the clinical point of view, the two techniques should be treated as two separate entities.