1. We have assessed the relationship between salt intake and 24 h ambulatory arterial pressure in middle aged men with essential hypertension. 2. During the run-in phase (1 month) we estimated the habitual sodium intake (the average Na excretion of two 24 h urine collections) of each participant (n = 14). In the randomized and crossover part of the study we contemplated a 'habitual' sodium intake phase, in which each individual received a fixed diet (about 30 mmol of Na+ and 65 mmol of K+) with additional salt so as to equalize the average intake of the run-in phase, as well as high sodium phases (habitual intake +50 and +100 mmol/day) and low sodium phases (habitual intake -50 and -100 mmol/day). After the trial, 10 patients underwent an additional week of fixed salt intake to assess the reproducibility of 24 h ambulatory monitoring. 3. Average 24 h arterial pressure at habitual sodium intake was significantly lower than that at high intake and significantly higher than at low sodium intake. Clinic arterial pressure showed similar trends but only systolic pressure changes at low sodium intake achieved statistical significance. 4. Analysis of the data on an individual basis showed a linear increase in 24 h mean arterial pressure with increasing levels of sodium intake in all but two cases (flat response in one case and a non-linear rise in the other case). The response pattern of clinic measurements was much less homogeneous. In the aggregate, there was a highly significant linear trend for ambulatory arterial pressure to rise with increasing levels of salt intake.(ABSTRACT TRUNCATED AT 250 WORDS)