In the presence of a Na+-gradient (out greater than in), L-glutamic acid and L-and D-aspartic acids were equally well concentrated inside the vesicles, while no transport above simple diffusion levels was seen by replacement of Na+ by K+. Equilibrium uptake values were found inversely proportional to the medium osmolarity, thus demonstrating uptake into an osmotically sensitive intravesicular space. The extrapolation of these lines to infinite medium osmolarity (zero space) showed only a small binding component in acidic amino-acid transport. When the same experiment was performed at saturating substrate concentrations, linear relationships extrapolating through the origin but showing smaller slope values were recorded, thus indicating that the binding component could be more important than suspected above. However, binding to the membrane was neglected in our studies as it was absent from initial rate measurements. Na+-dependent uphill transport of L-glutamic acid was stimulated by K+ present on the intravesicular side only but maximal stimulation was recorded under conditions of an outward K+-gradient (in greater than out). Quantitative and qualitative differences in the K+ effect were noted between pH 6.0 and 8.0. Initial uptake rates showed pH dependency in Na+-(out greater than in) + K+-(in greater than out) gradient conditions only with a physiological pH optimum between 7.0 and 7.5. It was also found that a pH-gradient (acidic outside) could stimulate both the Na+-gradient and the Na+ + K+-gradient-dependent transport of L-glutamic acid. However, pH- or K+-gradient alone were ineffective in stimulating uptake above simple diffusion level. Finally, it was found that increased rates of efflux were always observed with an acidic pH outside, whatever the conditions inside the vesicles. From these results, we propose a channel-type mechanism of L-glutamic acid transport in which Na+ and K+ effects are modulated by the surrounding pH. The model proposes a carrier with high or low affinity for Na+ in the protonated or unprotonated forms, respectively. We also propose that K+ binding occurs only to the unprotonated carrier and allows its fast recycling as compared to the free form of the carrier. Such a model would be maximally active and effective in the intestine in the in vivo physiological situations.