In recent studies of bilirubin encephalopathy, in situ flushing of the cerebral vessels has been used to clear blood from the brain. The effectiveness of such procedures has not been adequately documented. Herein young, male Sprague-Dawley rats were given about 750 KBq of 51Cr-labelled rat erythrocytes 3 min prior to sacrifice. There were four experimental groups: control, displacer (sulfisoxazole), hyperosmolality, and hypercarbia. Half of the rats in each group had the brain vasculature flushed in situ, while the remaining rat brains were not flushed. The brains were dissected into seven regions, and the radioactivity in the tissues was compared to that of blood drawn from the rats immediately before death. Significant amounts of blood (22-42%) remained after in situ flushing. Retention was significantly higher in the hyperosmolar animals, and significantly lower in the hypercarbic animals as compared to controls. Interregional differences in blood volumes per gram wet weight were significant without, but not with flushing. Similar observations were made using 125I-IgM as a marker for the plasma compartment. In studies of brain uptake of substances with high plasma concentrations, substance remaining within the cerebral vessels may contribute significantly to the apparent brain uptake values.