To examine the effects of reducing sodium intake upon the renin-angioten-sin-aldosterone system (RAAS), 5 healthy men and 5 healthy women, 17-37 years old, living under standardized conditions, were sampled around the clock, once on habitual and once on restricted sodium intake. Plasma renin activity (PRA), aldosterone (PA) and cortisol (PC) were determined by radioimmunoassay. All three variables were found to exhibit a statistically significant circadian rhythm, both on habitual and restricted salt intake. After salt restriction, an increase in midline-estimating statistic of rhythm (mesor) of PRA and PA, but not of PC, was observed. The acrophase (an estimate of the time of high values) for PC lagged behind that for PRA and PA. This difference in acrophase was of specially high statistical significance when subjects were on a sodium-restricted diet. These results demonstrate the importance of inferential statistical so-called rhythmometric methods: parameters such as the acrophase can also be used for the assessment of novel effects and for a quantification in time. The derivation of confidence intervals for each rhythm parameter allows one to verify that a given variable exhibits values bracketing an average not only between a higher and a lower, but also between an earlier and a later limit. Changes that may involve only the acrophase, such as a lead or lag, as here noted, are then detected and are of factual as well as methodological interest.