White coat hypertension (WCH) is common in referred hypertensive patients. Ambulatory blood pressure monitoring (ABPM) is not free from the white coat syndrome. We examined the use of the elevation of the first and last measurements of ABPM for diagnosis of WCH in a hypertensive population that had been referred to a hospital-based hypertension unit. Data were obtained on 1350 patients for clinic and ABPM parameters. WCH, as diagnosed by conventional clinic blood pressure (BP) measurement, was compared with a variety of alternative methods determined from ABPM. In all cases, mean daytime pressure was <135 mm Hg/85 mm Hg with an elevation of clinic BP >/=140 mm Hg systolic or 90 mm Hg diastolic. The definitions tested for this elevation were first hour mean pressure, first reading, maximum reading in first hour, last hour mean pressure, last reading, maximum reading in the last hour and maximum reading in first or last hour. Elevation of the maximum pressure in the first hour or last hour above 140 mm Hg systolic or 90 mm Hg diastolic showed a high level of agreement (kappa=0.91) with classical WCH for diagnosis of the white coat syndrome. Termed ambulatory white coat hypertension, patients with this finding were older than classic white coat patients and had higher daytime (127+/-6/78+/-5 mm Hg versus 121+/-5.5/74+/-6 mm Hg, P<0.005 for systolic and diastolic) and nighttime (114+/-11/67+/-8 mm Hg versus 106+/-9/61+/-6 mm Hg, P<0.005 for systolic and diastolic) pressures. They also had a significantly greater Sokolow-Lyon index (leads V(1)+V(5), 21+/-7 mV versus 18+/-6 mV). Elevation of BP above 140 mm Hg systolic or 90 mm Hg diastolic in the first or last hour of monitoring diagnoses patients with a white coat response in whom there is a higher BP profile than in patients with classic white coat response alone. We suggest, therefore, that this is a better measure of the white coat phenomenon.