The effects of an 8-h period of reduced colloid oncotic pressure (COP) on cerebral edema formation following cryogenic brain injury were studied in 24 normothermic, normocarbic rabbits. Anesthesia was induced with halothane and catheters inserted to permit the monitoring of arterial, right atrial, and intracranial pressures (MAP, CVP, and ICP, respectively). When surgery was complete, halothane was discontinued, 40 mg/kg of pentobarbital was given iv, and ventilation continued with 66% N2O/balance O2. A left parietal cryogenic injury was then produced using liquid N2, and the animals assigned to one of three groups. In group 1 (Control, n = 8), only maintenance lactated Ringer's solution (LR) was given for the remainder of the study. Beginning 30 min after injury, animals in the other two groups (n = 8 each) underwent 45 min of plasmapheresis (arterial phlebotomy, with packed cells returned). In group 2 (Iso-COP) separated plasma was replaced with 6% hetastarch in LR, while in group 3 (Hypo-COP), plasma was replaced with LR alone. In both groups, the volume of fluid given was adjusted to maintain MAP and CVP at baseline values. After plasmapheresis, animals subsequently recieved only maintenance LR. Eight and one-half hours after the injury (8 h after the start of plasmapheresis), animals were killed and the brains removed. Regional tissue specific gravities (SpGr) and water contents (%H2O) were measured respectively by microgrvimetry and drying. In addition, the %H2O of samples of skeletal muscle and small bowel were determined to assess peripheral edema formation. There were no important intergroup differences in MAP, CVP, ICP, blood gases, or osmolality at any time. In group 3, COP was reduced acutely from 20.6 mmHg to 10.0 mmHg, a value that remained stable for the remainder of the study. By contrast, COP in groups 1 and 2 decreased by only ~3 mmHg. In spite of these differences in COP, there were no intergroup differences in SpGr or %H2O between the groups. The %H2O of skeletal muscle and jejunum, however, was significantly higher in the Hypo-COP group. The authors conclude that while a prolonged reduction in COP leads to the development of peripheral edema, it has no effect on cerebral edema following an experimental brain injury.