In urban Australia, the risk of retiring with hypertension is greater than 40%, and the basic abnormality--a rise in blood pressure (BP) with age--is almost universal. A hypothesis linking this with salt, therefore, concerns everyone. The diet of early humans was unsalted, and the Na content of breast milk (6 mmol/kg) shows how little NaCl is needed even during the most rapid period of growth. The hypothesis that the hypertonic concentration needed to preserve food causes the BP to rise with age is based partly on the normotensive status of contemporary "salt-free" societies and partly on experimental evidence. "Salt-free" populations seldom use alcohol and happen to be lean and active, with a low fat intake and largely vegetarian diet, but Westerners with similar virtues do not escape hypertension. Ideally, the prophylactic effect of avoiding salt would be ascertained in large-scale, prospective trials, but practical, ethical and, economic factors impose serious design problems. Nevertheless, a public health intervention based on this hypothesis would be incomplete without a serious attempt to measure the outcome.