EFFORTS to explain the clinical course of patients receiving extensive superficial burns have led many clinicians to postulate the presence of a circulating toxin. Most of the experimental evidence presented in support of this view has been subject to the criticism that the treatment accorded the material removed from the burned animal might have contributed to its effect on the recipient. Since a simple transfusion of blood from a burned animal to a normal animal is generally well tolerated by the latter, it was decided to increase the possible sensitivity of the recipient to a hypothetic toxin by subjecting it, under ether anesthesia, to a standardized burn prior to the transfusion. The burn had been shown to result in a mortality of approximately 50 per cent in untreated animals. A second series of controls was employed, which was given transfusions of equivalent amounts of blood from normal animals.