Localized edema follows the freezing of a small area of cerebral cortex. Effects of five subsequent hours of anesthesia on this edema were studied in six groups of six dogs each. Six anesthetic techniques were studied. In six additional “awake” dogs, anesthesia (halothane) was discontinued immediately after the lesion was made. Eight control dogs received neither anesthesia nor cryogenic injury. Control white matter contained 67.4 = .4 (mean = SE) per cent water by weight. Twenty-four hours after the cryogenic injury water accounted for the following percentages of total weight of white matter adjacent to the lesion: 60 mg/kg pentobarbital. 73.2 = .9; 70 per cent N2O/Innovar, 73.6 = .9; “awake”. 77.9 = .9; 1.95 per cent enflurane. 78.2 = .9: 1.33 per cent iso-flurane. 78.6 = .8; 0.86 per cent halothane, 78.2 = .6; 1.89 per cent halothane, 79.7 = .6 Peak intracranial pressures (ICP) were 15.4 = 1.3 torr with pentobarbital. 21.6 = 1.8 torr with N2O /In-novar, and 31.1 = 2.6 to 38.3 = 45 torr with the halogenated anesthetics. The water content of white matter and ICP were significantly lower (P< 0.05) in animals receiving pentobarbital or N2O /Innovar anesthesia than in animals receiving inhalation anesthetics. The authors conclude that pentobarbital and fentanyl-droperidol (Innovar) limit the extent of cerebral edema, but that inhaled anesthetics do not.