Essential hypertension, a highly prevalent and devastating disease, likely begins in childhood, and presumably might respond at this stage to general preventive measures. The problems encountered in the recording of reproducible blood pressure levels are largely of a quantitative nature, and involve instrument validity, observer (examiner) training, and the interaction between the child and the physician's office environment. The problems of early diagnosis of hypertension are related to the lack of long-term observations of children. Population percentiles of normal blood pressure values could aid in the early diagnosis of hypertension. For this purpose the method of blood pressure measurement in the physician's office has to be strictly comparable to methods employed in accumulating the percentile or reference observations. Repeated measurements under basal-like conditions are advised, and the weight or height of the child is more important for standardization than is age.