The rates of anion net efflux from gramicidin-treated erythrocytes in the presence of a K gradient were measured at 25 degrees C, pH 7.8, as rates of loss of Ki. The experiments served to estimate the relative contributions of two hypothetical mechanisms to Cl net efflux at low extracellular Cl concentrations. Cl, Br, and NO3 net effluxes were measured into media of different Cl, Br, or NO3 concentrations, respectively, to determine and compare the relative rates of the extracellular anion-inhibitable components. They were 48, 160, and 230 mmol/(kg Hb X min), respectively, at a membrane potential of about -90 mV. This indicates that the anion-inhibitable efflux is not due solely to the return translocation of the empty transport site ("slippage") because slippage should be independent of the chemical nature of the anion. Cl net efflux was also measured as a function of the intracellular Cl concentration into media containing either 0 or 50 mM Cl. Under both conditions, net efflux was linearly dependent on Cli between 30 and 300 mM Cli and was 0 when back-extrapolated to 0 Cli. This observation is not compatible with the slippage process, which under these conditions would have been expected to be independent of Cli above 15 mM Cli. It was concluded that slippage contributes negligibly to Cl net efflux even at low extracellular anion concentrations and that the alternative process of "tunneling"--that is, movement of the anion through the anion transporter without a conformational change in a channel-type behavior--is the major, if not the sole, mechanism underlying Cl conductance.