Circadian rhythm of 18-hydroxycorticosterone (18-OHB), a possible immediate precursor of aldosterone (Fraser and Lantos 1978), has been demonstrated in man (Wilson, Mason and Fraser 1976; Sowers, Martin and Beck 1983) and rats (Imaizumi, Yamamoto, Kamei, Yoshida, Miyauchi, Kigoshi, Hosojima, Uchida and Morimoto 1987). This circadian rhythm of plasma 18-OHB resembles that of plasma aldosterone or glucocorticoid under regular conditions, whereas their circadian peaks occur in the morning in the diurnally-active species man and at night in the nocturnally-active ones rat. In man (Mason, Fraser, Morton, Semple and Wilson 1977; Sowers, Martin and Beck 1983), dietary sodium restriction causes a greater increase in plasma 18-OHB than in plasma aldosterone with a resultant increase in the plasma 18-OHB/aldosterone ratio. The circadian rhythm of plasma 18-OHB has also been reported to be similar under low and high sodium intakes (Sowers, Martin and Beck 1983). In the rat, there is no report on the effect of altered sodium intake on the circadian rhythm of plasma 18-OHB. The present study was performed in order to determine the circadian rhythms of plasma 18-OHB under low and high sodium intakes in conscious rats and to evaluate their correlations with those of plasma aldosterone or corticosterone.were available ad libitum for a recovery period of 3 days. Thereafter the cannulated rats were divided into 3 groups; control rats were maintained on a regular diet and tap water, sodium-depleted rats on a low sodium diet (Na 0.095 mg/g, K 4.8 mg/g) and tap water, and sodium-loaded rats on a regular diet and 1% saline to drink. Food and fluid intakes and urine volume were recorded daily. One week later, many of these rats were used for measurement of the circadian rhythms of plasma corticosteroids. Blood samples (0.3 ml) were collected from the cannulas into tubes at 6-h intervals starting 2000 h. Some were decapitated and trunk blood samples were immediately collected into tubes for measurements of plasma renin activity (PRA), electrolytes and hematocrit. These blood samples were centrifuged at 4°C and the plasmas were stored at -20°C until assayed. PRA was measured by direct RIA. Plasma corticosterone, 18-OHB and aldosterone were measured by a peak-height method and RIAs, respectively, as previously described (Imaizumi et al. 1987). Results are expressed as the mean±SEM. Statistical analysis was made by Student's t-test or analysis of variance.