Blood pressure assessment by a physician elicits an alerting reaction and a pressor response in the patient. The magnitude and time course of this response are described for a large number of hypertensive subjects in whom the assessments were performed during ambulatory intra-arterial blood pressure monitoring. In nearly all of the subjects, the physician's visit was accompanied by blood pressure and heart rate increases that peaked within 4 minutes and then declined. The response was characterized by a relatively high average value; a large between-subject variability; no relation with patient age, baseline hemodynamic values, and responses to laboratory stressors; and no attenuation with multiple repetition of the physician's visit. On the other hand, the increase in blood pressure was considerably less when blood pressure assessment was made by a nurse than when it was made by a physician; in both instances, a 10-minute wait was associated with marked reduction of the initial response. Thus, the stress inherent in usual blood pressure-measuring procedures is responsible for considerable overestimations of patients' blood pressures. There are means by which this can be minimized, although a residual error is likely to remain in most subjects. Whether the stress-devoid blood pressure is a better prognostic index than the stress-related one remains unknown.