Brain edema (BE) research lacks quantitative regional methods. We modified the method of Pappius and McCann (1969), who used radioactive iodinated I125 serum albumin (RISA) as a label for vasogenic BE fluid. To correct for intravascular plasma volume we used Cr51 labelled red blood cells and calculated equivalent extravascular plasma volume (EVPV). The modified RISA method was compared with a standard method for measuring increased tissue water, i.e. the change in wef.dry weights. Anesthetized rabbits were subjected to unilateral cortical freeze injury and sacrificed three hours later. The lesion corresponding to the area of blood brain barrier (BBB) breakdown was delineated by Evan's blue staining. That area and the two adjacent poles were sectioned. The contralateral hemisphere was used as the control. Good agreement was found between the two methods. By the modified RISA method, 4.08% of the wet weight of the injured hemisphere was EVPV while the decrease in dry weight corresponded to 5.56% edema. In the freeze lesion, however, where BBB breakdown occurred extravascular (EV)protein exceeded EV water, while adjacent to the lesion in the area with intact BBB, EV water exceeded EV protein. This suggests that the diffusion rate for water through brain tissue exceeds that of albumin in the early stages of vasogenic BE formation. We conclude that the modified RISA method is a satisfactory measure of regional vasogenic BE in acute animal experiments.