1. A first sample of a cross-section of the Belgian population was studied.
2. The presence of hypertension or borderline hypertension was established on the basis of two examinations at 1–3 weeks intervals, during which the blood pressure was measured four times after 10 min rest.
3. The prevalence of both hypertension and borderline hypertension increased with age in the two sexes and, independently of age, with the relative weight.
4. The probable effect of arm circumference in overestimation of blood pressure in heavy subjects was analysed by multivariate correlations. It was concluded that if the influence of arm circumference is excluded, the effect of relative weight on blood pressure is less marked but still present.