Glycerol was administered to control rats and to animals with experimental cerebral microembolism as a single intravenous injection (1 gm per kilogram), a single one-hour intravenous infusion (1.5 gm per kilogram), daily one-hour intravenous infusions (1 gm per kilogram), or in multiple oral doses (1 gm per kilogram per four hours). There were no effects of glycerol on the mortality, brain edema or increased brain sodium concentration and75Seselenate space due to cerebral microembolism. Brain water was not reduced in the hemispheres contralateral to embolization or in glycerol-treated controls. The results suggest that glycerol, at the dose levels reported to be beneficial in human cerebral infarction, is ineffective for cerebral dehydration.