There is evidence of crucial central nervous system involvement in the pathogenesis of mineralocorticoid-excess salt hypertension, as well as data indicating that corticosterone is the predominant ligand for the type I adrenocorticoid receptor in the brain. Miniosmotic pumps were used to deliver artificial cerebrospinal fluid (CSF), aldosterone (10 ng/h), corticosterone (10 or 20 ng/h), aldosterone (10 ng/h) plus corticosterone [10 ng/h intracerebroventricularly (icv)], or aldosterone (10 ng/h) plus corticosterone (20 ng/h icv). All animals were sensitized to mineralocorticoid hypertension by removing the right kidney and offering saline to drink. Indirect blood pressure by the unheated tail-cuff method and weights were measured twice weekly; 24-h urine volumes were measured once a week. The blood pressures of the four groups did not differ statistically before infusion. The blood pressures of those animals receiving CSF or corticosterone were not significantly elevated after 4-5 wk of intracerebroventricular infusion, whereas the aldosterone group had become significantly elevated within 2 wk. A similar study was done comparing the effects of intracerebroventricular infusion of aldosterone (10 ng/h), aldosterone (10 ng/h) and RU26988 (20 ng/h), and RU26988 (20 ng/h). RU26988, a selective type II receptor agonist, had no effect on the blood pressure, nor did it alter the pressor effect of intracerebroventricular aldosterone. The concomitant infusion of corticosterone antagonized the increase in blood pressure in a dose-dependent manner. Neither steroid nor their combinations produced significant differences in daily urine volume or body weight gain compared with the CSF group.