Release of excitatory amino acids and dopamine plays a central role in neuronal damage after cerebral ischaemia. In the present study, we used an in vitro model of ischaemia to investigate the effects of sevoflurane on dopamine, glutamate and aspartate efflux from rat corticostriatal slices. Slices were superfused with artificial cerebrospinal fluid at 34 degrees C and episodes of 'ischaemia' were mimicked by removal of oxygen and reduction in glucose concentration from 4 to 2 mmol litre(-1) for < or = 30 min. Dopamine efflux was monitored in situ by voltammetry while glutamate and aspartate concentrations in samples of the superfusate were measured by HPLC with fluorescence detection. Neurotransmitter outflow from slices was measured in the absence or presence of sevoflurane (4%). After induction of ischaemia in control slices, there was a mean (SEM) delay of 166 (7) s (n = 5) before sudden efflux of dopamine which reached a maximum extracellular concentration of 77.0 (15.2) micromol litre(-1). Sevoflurane (4%) reduced the rate of dopamine efflux during ischaemia (6.90 (1.5) and 4.73 (1.76) micromol litre(-1) s(-1) in controls and sevoflurane-treated slices, respectively; P<0.05), without affecting its onset or magnitude. Excitatory amino acid efflux was much slower. lschaemia-induced glutamate efflux had not reached maximum after 30 min of ischaemia. Basal (pre-ischaemic) glutamate and aspartate efflux per slice was 94.8 (24.8) and 69.3 (31.5) nmol litre(-1) superfusate (n = 4) and was not significantly reduced by 4% sevoflurane. lschaemia greatly increased glutamate and aspartate efflux (to a maximum of 919 (244)% and 974 (489)% of control, respectively). However, ischaemia-induced efflux of both glutamate and aspartate was significantly reduced by 4% sevoflurane (P < 0.001 for glutamate, P < 0.01 for aspartate). In summary, sevoflurane may owe part of its reported neuroprotective effect to a reduction of ischaemia-induced efflux of excitatory amino acids and, to a lesser extent, dopamine.