During the course of a study of the immediate effect of transfusion on the blood count, certain data have appeared which seem to indicate that, besides the change in volume due directly to the transfusion, changes in the total volume occur, due to increase or decrease in the recipient's plasma. These apparent changes in the blood volume may take place within a short time after transfusion without inconvenience to the patient, and may be either increases or decreases. In certain patients, these changes do not occur, especially in those whose blood is comparatively normal. If they really are blood volume changes, they would be a source of error to any method of determining blood volume, which is dependent on the injection of a known amount of some constituent already present in the blood, and the subsequent measurement of the increase in that constituent, because they would vitiate the initial reading.