The clinical value of a set of three clinical blood pressure measurements as a predictor of daytime ambulatory hypertension was assessed by performing a set of clinical blood pressure measurements in 171 borderline hypertensives, and calculating their diagnostic accuracy, sensitivity, specificity and predictive value compared to the daytime average of ambulatory blood pressure monitoring. Diagnostic accuracy was 0.63, sensitivity was 81% and specificity was 47%. Positive and negative predictive values were 0.60 and 0.74, respectively. The set of clinical measurements detected 81% of hypertensives, but 36% of the population screened was mislabelled--48 patients (28%) as hypertensive and 15 (8%) as normotensive. A single set of clinic blood pressure measurements is quite sensitive for diagnosing daytime hypertension, although its accuracy, specificity and predictive value are low. The subpopulation incorrectly labelled as normotensive may have a different prognosis and merits further prospective study.