The objective of this prospective study was to define the limits below which ambulatory blood pressure (BP) does not decrease in patients with essential hypertension, when the decision to institute and intensify drug treatment is based on conventional blood pressure measurements. After a 1 month placebo run-in period, 30 patients were treated for 1 year with the converting enzyme inhibitor lisinopril or the calcium antagonist isradipine; dose adjustments and the decision to add hydrochlorothiazide were based on conventional blood pressure measurements in the clinic. Ambulatory blood pressure was recorded during 24 h in the run-in period and after 16, 24, and 52 weeks of active therapy. The baseline ambulatory blood pressure below which pressure does, on average, not decrease during active treatment was defined as the pressure at which the regression line between the on-treatment pressure and blood pressure in the run-in period intersects the line of identity. The systolic/diastolic blood pressure limits were similar for the three assessments during active treatment and averaged 128/88 mm Hg for daytime, 106/73 mm Hg for nighttime pressure, and 119/81 mm Hg for the whole 24 h, with upper 95% confidence limits of 137/93, 115/78, and 127/86 mm Hg, respectively. In conclusion, in the present study population, ambulatory diastolic blood pressure did not decrease in response to the antihypertensive therapy, which was guided by conventional blood pressure measurements, when its initial values were less-than-or-equal-to 88 mm Hg (upper limit: 93 mm Hg) for the day, less-than-or-equal-to 73 mm Hg (78 mm Hg) for the night, and less-than-or-equal-to 81 mm Hg (86 mm Hg) for the whole 24 h period.