1. Applicability of the previously postulated active transport system for conveying glucose into the human red cell was tested in connection with a number of related substances, comprising 6-carbon aldoses and ketoses, 5-carbon aldoses, and 5- or 6-carbon polyhydric alcohols.
2. The alcohols did not perceptibly penetrate the cells; all the sugars penetrated, the rates differing, but all of the same order of magnitude.
3. All the aldoses penetrated according to the pattern previously reported for glucose, in that the rate of penetration was not directly related to the gradient, but subject to some limiting factor.
4. The ketoses penetrated approximately according to the pattern of passive diffusion.
5. When present in equal concentrations, any aldose prevented or greatly delayed the entrance of a ketose, while the ketose did not perceptibly alter the rate of aldose uptake. Within each class, similar inhibitory relations were observed.
6. Penetration of all the sugars showed a high temperature coefficient and was inhibited by the mercuric ion or p-chloromercuribenzoate; certain of the sugars showed a residual degree of penetration not thus inhibitable.
7. Penetration of the ketoses was selectively inhibited by barbital.
8. These observations are interpreted in terms of simple equilibria between the various sugars and a hypothetical carrier molecule in the membrane, with which the sugars form a complex during their passage through the membrane. Comparisons between the sugars in relation to their reactions with the carrier system are indicated.