In earlier experiments in this laboratory it has been observed that dextran, fed to human subjects, is not recovered in the stools. Engstrom and Aberg have suggested that some of the dextran administered intravenously might be excreted into the gastrointestinal tract where it could be destroyed by bacterial action. Our own observations on dextran incubated with stools suggested that dextran disappears only very slowly, but Hehre has since shown that there are large numbers of anaerobic bacteria present in the intestinal flora, and that it is these, rather than the aerobes, that are capable of splitting dextran. Although ingested dextran may be in part degraded and consumed by the intestinal bacteria, it still seemed to us of interest to learn more about the disappearance of dextran fed by mouth. The following experiments show that dextran feeding leads to a sustained increase in blood sugar and a rise in liver glycogen in the fasted rat, and brings about an increase in the blood sugar of human subjects. Experimental. Male albino rats of the Sprague-Dawley strain, weighing 250-350 g were fasted for 24 hours. A group of 10 control animals were killed and both fractions of liver glycogen were determined by the method of Bloom, Lewis and Schumpert. Glycogen was determined (as glucose equivalent) by means of the anthrone reaction. The fasted experimental animals were given 5 ml of 18% dextran in 0.9% saline by stomach tube. An initial blood sample was taken from the cut end of the tail and additional samples were taken at intervals after feeding. Copper tungstate filtrates were prepared and reducing substances were determined by the method of Nelson. In a series of four rats, both total and fermentable reducing substances were determined. Two or four hours after feeding, the rats were anesthetized with Nembutal and the liver was removed and immediately frozen and powdered. Trichloroacetic acid-extractable and total glycogen was determined on aliquots of the well-mixed liver powder. In order to determine whether glycogen or dextran might be present, other samples were allowed to stand at room temperature for 1.5 to 2 hours in order that glycogenolysis might occur. Dextran was added to parallel samples in order to determine whether dextranolysis could occur in the liver preparation.