BACKGROUND: The difference between clinic and ambulatory average daytime blood pressures is frequently taken as a surrogate measure of the 'white-coat effect' (i.e. the pressor reaction triggered in the patient by the physician's visit). OBJECTIVE: To assess the reproducibility of this difference and its relationship with clinic and average ambulatory daytime blood pressure levels. DESIGN AND METHODS: These issues were addressed with two large groups of subjects in whom both clinic and ambulatory blood pressures were measured, namely 783 outpatients with systolic and diastolic essential hypertension [Group 1, aged 50.8+/-9.4 years (mean +/- SD)], participating in standardized Italian trials of antihypertensive drugs, and 506 elderly patients (group 2, age 71+/-7 years) with isolated systolic hypertension, participating in the European Syst-Eur trial. RESULTS: The clinic-daytime blood pressure difference for the essential systolic and diastolic hypertensive patients (group 1) was 13.6+/-14.3 mmHg for systolic and 9.1+/-8.6 mmHg for diastolic blood pressure (P always < 0.01). This difference for the elderly patients with isolated systolic hypertension (group 2) was 21.2+/-16.0 mmHg for systolic and only 1.3+/-10.2 mmHg for diastolic blood pressure (P < 0.01 and P < 0.05, respectively). In both studies little or no systematic clinic-daytime difference could be observed for heart rate. The reproducibility of the clinic-daytime blood pressure difference, tested for 108 essential systolic and diastolic hypertensive patients from group 1 and 128 isolated systolic hypertensives from group 2, was invariably lower than that both of daytime and of clinic blood pressure values. Finally, the clinic-daytime blood pressure difference was progressively higher for increasing levels of clinic blood pressure and progressively lower for higher levels of ambulatory daytime blood pressure. CONCLUSIONS: Thus, the clinic-daytime blood pressure difference has a limited reproducibility; depends not only on clinic but also on daytime average blood pressure, which means that its size is a function of the blood pressure criteria employed for selection of the patients in a trial; and is never associated with a systematic clinic-daytime difference in heart rate, which further questions its use as a reliable surrogate measure of the true pressor response induced in the patient by the doctor's visit.