&NA;This experiment was designed to study the effects of intravascular volume expansion with crystalloid and colloid agents on intracranial pressure (ICP) and computerized electroencephalographic (EEG) data in dogs with cerebral vasogenic edema. A liquid nitrogen cold lesion was produced in the right parietal lobes of 30 anesthetized mongrel dogs. ICP data from an intraventricular catheter, hemodynamic data, and EEG power data were recorded continuously during the course of the experiment. Six hours after creation of the liquid nitrogen cold lesion, each dog was infused with one of the following treatments: 1) no treatment (control); 2) 5% dextrose in water (D5W) (30 ml/kg/h); 3) 0.9% NaCI (30 ml/kg/h): 4) hetastarch (10 ml/kg/h); and 5) hetastarch (30 ml/kg/h). The animals were treated for 2 hours, killed, and the brains examined and measured for water content. During the D5W infusion, ICP increased significantly (141%) but the EEG power data did not change. During the normal saline infusion, ICP increased significantly (91%) and the EEG power data showed marked deterioration. Both hetastarch infusions produced no change in ICP and demonstrated an improvement in the EEG power data (P< 0.001). These data suggest that crystalloid infusions may have profound deleterious effects in the presence of brain edema, but colloids do not, and should be considered in the fluid management of patients with cerebral edema. (Neurosurgery25:173‐179, 1989)