Stimulus evoked neurotransmitter release requires that Na+ channel-dependent nerve terminal depolarization be transduced into synaptic vesicle exocytosis. Inhaled anesthetics block presynaptic Na+ channels and selectively inhibit glutamate over GABA release from isolated nerve terminals, indicating mechanistic differences between excitatory and inhibitory transmitter release. We compared the effects of isoflurane on depolarization-evoked [3H]glutamate and [14C]GABA release from isolated nerve terminals prepared from four regions of rat CNS evoked by 4-aminopyridine (4AP), veratridine (VTD), or elevated K+. These mechanistically distinct secretegogues distinguished between Na+ channel- and/or Ca2+ channel-mediated presynaptic effects. Isoflurane completely inhibited total 4AP-evoked glutamate release (IC50=0.42 ± 0.03 mM) more potently than GABA release (IC50=0.56 ± 0.02 mM) from cerebral cortex (1.3-fold greater potency), hippocampus and striatum, but inhibited glutamate and GABA release from spinal cord terminals equipotently. Na+ channel-specific VTD-evoked glutamate release from cortex was also significantly more sensitive to inhibition by isoflurane than was GABA release. Na+ channel-independent K+-evoked release was insensitive to isoflurane at clinical concentrations in all four regions, consistent with a target upstream of Ca2+ entry. Isoflurane inhibited Na+ channel-mediated (tetrodotoxin-sensitive) 4AP-evoked glutamate release (IC50=0.30 ± 0.03 mM) more potently than GABA release (IC50=0.67 ± 0.04 mM) from cortex (2.2-fold greater potency). The magnitude of inhibition of Na+ channel-mediated 4AP-evoked release by a single clinical concentration of isoflurane (0.35 mM) varied by region and transmitter: Inhibition of glutamate release from spinal cord was greater than from the three brain regions and greater than GABA release for each CNS region. These findings indicate that isoflurane selectively inhibits glutamate release compared to GABA release via Na+ channel-mediated transduction in the four CNS regions tested, and that differences in presynaptic Na+ channel involvement determine differences in anesthetic pharmacology.