In the present study, the effect of selective glucocorticoid deficiency on renal water excretion was investigated in conscious, trained, adrenalectomized dogs. The animals were studied before and after a water load while on replacement therapy of desoxycorticosterone acetate, 5 mg/day, and dexamethasone, 0.8 mg/day (group I), and while off dexamethasone for 5-9 days (group II). Before the water load the weight, inulin space, cardiac output, blood pressure, glomerular filtration rate, renal blood flow, plasma osmolality, and plasma antidiuretic hormone measured by radioimmunoassay were similar in both groups I and II. However, after a 40 ml/kg water load a marked impairment in renal water excretion in the glucocorticoid deficient dogs became apparent. Maximal free water clearance was -0.046+/-0.16 vs. 6.51+/-0.72 ml/min (P < 0.001) and minimal urinary osmolality was 425+/-56 vs. 82+/-3.5 mosmol/kg H(2)O (P < 0.001) in group II as compared to group I. Plasma antidiuretic hormone was maximally suppressed during the water load in group I to 0.34+/-0.08 pg/ml but remained elevated at 9.18+/-1.79 pg/ml (P < 0.005) in group II. This nonsuppressibility of plasma antidiuretic hormone during water loading in group II was associated with a significant tachycardia of 145+/-6 vs. 87+/-6 beats/min (P < 0.001) in group I and a significantly lower stroke volume of 27+/-0 vs. 59+/-0.5 ml/beat (P < 0.001). In conclusion, our results implicate a persistent secretion of antidiuretic hormone as an important factor in the impaired water excretion of glucocorticoid deficiency. A deleterious effect of glucocorticoid deficiency on cardiac function was observed and this hemodynamic alteration could be involved in initiating a nonosmolar, baroreceptor-mediated release of vasopressin.