This work deals with the possibility that cations may be transported through the human red blood cell membrane as negatively charged ion pairs of the type XCO3‐. Passive sodium influx into human erythrocytes was increased 4‐fold when 150 mM chloride was substituted with bicarbonate in the medium. This increased permeability to sodium was completely abolished when the cells were pretreated with a specific inhibitor of anion transport (DIDS). Both bicarbonate‐induced sodmm influx and chloride exchange were linearly related to DIDS‐binding. Both transport processes were completely inhibited by the binding of 1.1X 106DIDS molecules per cell. The presence of bicarbonate and/or pre‐treatment of the cells with DIDS did not change the permeability of the erythrocyte membrane to K+, Rb+, and Cs+to any significant extent. The experimental findings support the hypothesis that Li+and Na+, but not K+, Rb+, and Cs+form monovalent negatively charged ion pairs with CO3‐, which traverse the membrane through the anion exch