In INTERSALT, the relationship between body mass index (kg/m2) and blood pressure was studied in 10,079 men and women aged 20-59, sampled from 52 centres around the world, based on a standardised protocol with central training of observers, a central laboratory and extensive quality control. Body mass index-blood pressure relationships were first studied in men and women within each centre, and results of these regression analyses were then pooled for all 52 centres. With adjustment for age, alcohol intake, smoking, and sodium and potassium excretion, body mass index was positively associated with systolic blood pressure among men in 51 of 52 centres and among women in 47, significantly so in 24 and 27, respectively. Body mass index was positively associated with diastolic blood pressure in 51 and 49 centres in men and women, respectively, significantly so in 33 and 31. When the within centre regression coefficients were pooled, the pooled coefficients were highly significant in all analyses (P less than 0.001), and were similar for those aged 20-39 and 40-59. Overall, a 10 kg difference in body weight was associated on average with a 3.0 mmHg difference in systolic and a 2.2 mmHg difference in diastolic pressure. In further analyses across centres, median body mass index was related significantly to median systolic blood pressure, median diastolic pressure and the prevalence of hypertension in both men and women. Body mass index was related to the slopes of systolic and diastolic blood pressure with age in women, but not in men. These INTERSALT findings confirm the importance of the association between body weight and blood pressure. Differences of 2-3 mmHg in systolic blood pressure on a population basis have been shown to be associated with differences in stroke mortality rates of 6-9 per cent and in coronary death rates of 4-6 per cent.