The contributions of sodium and chloride ions and of osmotic stresses to the ischemia-evoked efflux of excitotoxic and other amino acids were explored using a rat four vessel occlusion model. Replacement of Na+ with choline or N-methyl-D-glucamine (NMDG) and of Cl- with sulfate or gluconate was used to evaluate the contribution that these ions make to amino acid efflux. The contribution of ischemia-evoked swelling to amino acid release was studied by applying mannitol or sucrose to minimize the cell volume increases and the compensatory regulatory volume decrease evoked efflux of amino acids. Aliquots of artificial cerebrospinal fluid (aCSF), appropriately adjusted for ion replacement or 150 mM mannitol or sucrose, were pipetted into cortical cups 35 min prior to ischemia and perfusate samples were obtained prior to, during and following ischemia (20 min) and reperfusion (40 min). Replacement of Na+ by NMDG depressed basal (normoxic) efflux of most amino acids, with choline substitution having little effect. During ischemia NMDG substitution increased glutamate and GABA efflux and choline enhanced the release of aspartate, glutamate, GABA and taurine. A reduction in extracellular Na+ would facilitate reversal of Na+-dependent transporters with extrusion of amino acids. Another possible explanation for the elevated release is that the absence of Na+ would inhibit the Ca2+/Na+ counter transport system, with a deleterious accumulation of intracellular Ca2+. Chloride replacement with sulfate or gluconate enhanced the efflux of aspartate, glutamate, GABA and taurine during ischemia. Removal of Cl- would depolarize cells, and block the Cl--dependent action of inhibitory amino acid transmitters, with both actions enhancing the ischemic injury and, consequently, amino acid release. Exposure to hyperosmotic mannitol (150 mM) aCSF enhanced ischemia-evoked release of some amino acids (taurine, GABA) and decreased that of aspartate and phosphoethanolamine. Sucrose aCSF enhanced the ischemia-evoked release of most amino acids. A potential explanation for these observations is that both agents may be able to rapidly penetrate the plasma membranes of ischemic neurons, actually contributing to the release of other osmolytes. The unanticipated nature of many of the observations made during these manipulations of the aCSF serves to accentuate the complex nature of the mechanisms responsible for the ischemia-evoked amino acid efflux into the interstitial spaces.