The intraventricular administration of vasopressin or DDAVP (desmopressin acetate) increased the brain water content from 78.2% to 79.2-79.5%. This was achieved without an accompanying water load. The applied water load alone did not increase the water content of the brain. There was no significant difference in the water content of the brain between animals treated with intraventricular vasopressin and intravenous water load and animals receiving only intraventricular vasopressin. The water content of the olfactory bulbs of the control animals was 3.8% higher than that of the hemispheres. While the water content of the hemispheres increased by 1.3%, that of the olfactory bulbs did so by 1.7% subsequent to the intraventricular administration of DDAVP. Measurement of the brain electrolyte content was not conclusive as to the mechanism of water permeability changes. The possible mechanism is discussed. Although no tissue or cerebrospinal fluid concentrations of vasopressin enabling comparison with clinical pathological conditions have been measured, it is suggested that increased secretion of vasopressin into the cerebrospinal fluid in conditions such as subarachnoid hemorrhage or intracranial hypertension of various origins might play a role in edema formation.