The early detection and treatment of primary and secondary hypertension in children may contribute to the improvement of health later in life. During childhood, the occurrence of arterial hypertension should be understood as a risk situation, even if the clinical signs of the disease are not obvious. Comparatively to other populations, only few epidemiological studies exist in Portugal to be used as clinical references. Thus, the aim of this work was to study the distribution curves of 30, 60 and 90 percentiles of systolic blood pressure (SBP) and diastolic blood pressure (DBP), related to age, height, and weight. To define the "cut off points" of hypertension in our population, we estimated the 95 percentile of SBP and DBP. This is a cross-sectional study. The sample consisted of 474 children (242 boys and 232 girls), aged 8 to 13 years old, randomly selected in 30 different schools of the Oporto region. Weight, height and body mass index were evaluated as anthropometric reference values and blood pressures, SBP and DBP, were measured at rest. The results revealed a gradual increase in blood pressure values with age, height and weight. We found significant correlations (p < or = 0.05; p < or = 0.01) between SBP and DBP to the variables height, weight and blood mass index in both sexes. Multiple regression analysis showed that weight and age explain significantly (p < or = 0.01) the variation of SBP in both sexes. Related to DBP, in males and females, only age accounted significantly for the amount of variance. In conclusion, the present study pointed out the need to associate the SBP and DBP values to the age, weight and height of the subjects, mainly in boys. This work stresses the need for SBP and DBP percentile tables related to age, to weight and to height, for the Portuguese population, particularly for pediatric ages.