ObjectiveTo determine whether the prediction of target-organ damage varies according to the reproducibility of 24 h blood pressure.SettingSeventeen hypertension clinics in northeast Italy.Main outcome measuresCorrelations of left ventricular mass index and albumin excretion rate with 24 h and office blood pressures in relation to tertiles of ambulatory blood pressure reproducibility.Patients and methodsIn 716 consecutive, stage I, hypertensives enrolled in the Hypertension and Ambulatory Recording Venetia Study (HARVEST), ambulatory blood pressure monitoring was performed twice, 3 months apart. In all subjects, the albumin excretion rate was measured by radioimmunoassay, and in 567, the left ventricular mass index was assessed by echocardiography.ResultsThe subjects were divided into tertiles of ambulatory blood pressure consistency (between-monitoring differences, regardless of the sign). In the tertile of subjects with good reproducibility, correlation coefficients of systolic and diastolic ambulatory blood pressure with left ventricular mass and urinary albumin excretion were significant and higher than those of office blood pressure. In contrast, in the two tertiles with poorer reproducibility, the coefficients were barely or not significant for both pressures. The advantage of ambulatory blood pressure over office blood pressure in predicting target-organ damage was no longer present for systolic blood pressure differences greater than 3.8 mmHg and diastolic blood pressure differences greater than 3.1 mmHg.ConclusionsThese data indicate that ambulatory blood pressure is a better predictor of left ventricular mass and urinary albumin excretion than office blood pressure, but only in subjects with good pressure reproducibility. Therefore, the assessment of hypertensive patients should be based on duplicate blood pressure monitorings. Recordings with 24 h systolic and diastolic blood presssure differences greater than 4 and 3 mmHg, respectively, should be considered with caution.