Index subjects were selected from the population of Milton studied in the May 1981 survey. Index subjects aged 64 or less with a systolic blood pressure 138-179 mmHg (18.35/23.8 kPa), including those on antihypertensive treatment, were invited to participate with their families. Index subjects were divided into two matched groups. One group was randomly assigned to be the control group and the other to be the salt-restriction group. Age, weight, height and 24-hour urinary excretion of sodium, potassium and creatinine at the start of the study in August 1981 were similar for the control and salt-restriction groups. When tested in November 1981, the salt-restriction group had achieved a reduction in 24-hour sodium excretion to a mean of 84 mmol for men and 70 mmol for women; corresponding values for the control groups were 150 and 120 mmol. A further test in March 1982 showed little further change. Potassium output changed very little. Attitudes to the low-salt diet varied, but 87 percent found it tolerable or actually preferable. There is no doubt that major reductions in sodium intake are feasible. However, if these are to be achieved on a large scale, food, manufacturers will need to offer a variety of low-sodium foods.