In 1931 Wierzuchowski (1) showed an influence of insulin on galactose tolerance of dogs. More recently, Levine, Goldstein, Huddlestun, Klein, and Henry (2, 3) demonstrated that insulin lowers the blood levels of D-galactose, D-xylose, and L-arabinose but not of D-arabinose or D-mannose, following their infusion into the eviscerated nephrectomized dog. The volume of distribution of the responsive sugars was increased from 45 to 70 per cent of body weight as a result of insulin administration. It was suggested that the insulin response represented a facilitation of transport of the sugar across the cell membrane into intracellular fluid. In all the responsive sugars the stereochemical structure of carbon atoms 1 to 3 was like that of D-glucose, and these authors suggested that these structural characteristics determined the insulin responsiveness of a sugar.Confirmatory evidence has appeared that insulin affects transport of D-galactose, D-xylose and L-arabinose across the cell membrane in the rat diaphragm (4-6), eviscerate rat (7), perfused rat heart (8) and intact cat (9). However, slow penetration of a number of sugars into muscle of nephrectomized and eviscerated rats has now also been demonstrated, and this slow penetration is increased by insulin for D-lyxose, D-ribose, D-mannose, D-fructose and D-fucose, as well as for D-xylose, D-galactose and L-arabinose (10, 11). Although these results indicate that the concept of structural specificity in carbon atoms 1 to 3 for insulin responsiveness has proved to be too limited in scope, the lack of an effect of insulin upon intracellular entry of sorbitol, mannitol and raffinose is strong evidence that the action of insulin is not simply an acceleration of a process of physical diffusion (12).