BACKGROUND: Spinal cord injury occurring as a result of surgical repair of thoracic and thoracoabdominal aortic disease remains a devastating complication. Excitatory amino acids have been known to cause neurologic injury after neuronal ischemia. The purpose of this study was to elucidate the effects of intrathecal bupivacaine or bupivacaine dextrose on neurologic outcome and to characterize NMDA receptor gene expression in transient spinal ischemia. METHODS: Rats were anesthetized with isoflurane, divided by 3 groups: Control (C group), Intrathecal 0.5% bupivacaine 40 microliter (B group) and Intrathecal 0.5% bupivacaine with 8.5% dextrose 40 microliter (D group). Spinal ischemia was produced by induced hypotension and thoracic aortic cross clamping. After spinal ischemia, neurologic scores were assessed at 1, 2, 3, 24, and 48 hours. After 3 hours rats (Sham, C, B, and D groups) were euthenized and spinal cords and cerebral cortexes were removed for to assay NMDAR mRNA. 48 hours after ischemia, rats (C and B groups) were euthenized and spinal cords were removed for histologic examination. RESULTS: The neurologic scores of B group were significantly lower than those of C group or D group. Rats of D group showed seizure 1 hour after ischemia and died about 3 hours after ischemia. There were no significant changes in the NMDAR mRNA expressions of the cerebral cortex and the spinal cord. C group showed more significant apoptosis by TUNEL staining than B group. CONCLUSIONS: Intrathecal bupivacaine, not bupivacaine dextrose, proved effective at preventing neurologic injury after transient spinal ischemia. Intrathecal dextrose might be involved in the induction of generalized tonic-clonic seizure after spinal ischemia.