The role of presynaptic mechanisms in general anesthetic depression of excitatory glutamatergic neurotransmission and facilitation of GABA-mediated inhibitory neurotransmission is unclear. A dual isotope method allowed simultaneous comparisons of the effects of a representative volatile (isoflurane) and intravenous (propofol) anesthetic on the release of glutamate and GABA from isolated rat cerebrocortical nerve terminals (synaptosomes). Synaptosomes were prelabeled with L-[(3)H]glutamate and [(14)C]GABA, and release was determined by superfusion with pulses of 30 mM K(+) or 1 mM 4-aminopyridine (4AP) in the absence or presence of 1.9 mM free Ca(2+). Isoflurane maximally inhibited Ca(2+)-dependent 4AP-evoked L-[(3)H]glutamate release (99 +/- 8% inhibition) to a greater extent than [(14)C]GABA release (74 +/- 6% inhibition; P = 0.023). Greater inhibition of L-[(3)H]glutamate versus [(14)C]GABA release was also observed for the Na(+) channel antagonists tetrodotoxin (99 +/- 4 versus 63 +/- 5% inhibition; P < 0.001) and riluzole (84 +/- 5 versus 52 +/- 12% inhibition; P = 0.041). Propofol did not differ in its maximum inhibition of Ca(2+)-dependent 4AP-evoked L-[(3)H]glutamate release (76 +/- 12% inhibition) compared with [(14)C]GABA (84 +/- 31% inhibition; P = 0.99) release. Neither isoflurane (1 mM) nor propofol (15 microM) affected K(+)-evoked release, consistent with a molecular target upstream of the synaptic vesicle exocytotic machinery or voltage-gated Ca(2+) channels coupled to transmitter release. These findings support selective presynaptic depression of excitatory versus inhibitory neurotransmission by clinical concentrations of isoflurane, probably as a result of Na(+) channel blockade.