Standardised data on blood pressure, 24 h urinary electrolyte excretion, body mass index (BMI) and alcohol intake were collected as part of the INTERSALT study in 598 men and women aged 20-59 years, selected randomly from three population groups in the United Kingdom. For the three centres combined, mean systolic blood pressure was 121.4 mm Hg and diastolic pressure 72.1 mm Hg, urinary sodium excretion 152.1 mmol/24 h, urinary potassium excretion 61.0 mmol/24 h, urinary sodium/potassium ratio 2.64 and BMI 25.2 kg/m2. Prevalence of heavy alcohol drinking in men (greater than or equal to 300 ml/week) was 27.5 per cent. Applying overall INTERSALT regression coefficients to the United Kingdom data suggested that modest changes in average sodium and potassium intakes, together with reductions in the prevalence of obesity and (in men) of heavy alcohol drinking could lead to important reductions in average population blood pressures and the prevalence of hypertension. The potential of this multifactorial approach to blood pressure control was illustrated by stratifying individuals within each of the United Kingdom centres by sodium and potassium excretion, BMI and alcohol intake. The 20 (out of 299) men considered at 'lower risk' for high blood pressure with respect to the above variables had systolic pressure lower by 11 mm Hg (P less than 0.01); for the 27 (out of 299) 'lower risk' women, systolic pressure was lower by 5 mm Hg (P = 0.06). These non-pharmacological approaches towards more favourable blood pressure levels could be accompanied by reductions in mortality from stroke and coronary heart disease.