There is a considerable difference in the number of reports of neurologic injury in the literature between lidocaine and other local anesthetics. Few in vivo animal studies have produced convincing results showing a difference in neurotoxicity among anesthetics. We investigated whether lidocaine and bupivacaine differ with respect to sensory impairment and histologic damage when equipotent doses of the two are administered intrathecally in rats. First, to determine relative anesthetic potency, rats intrathecally received 20 mu L of saline, 0.625%,1.25%,2.5%, or 50% lidocaine, or 0.125%, 0.25%, 0.5%, or 1.0% bupivacaine, and were examined with the tail-flick test for 90 min. The potency ratio calculated was approximately 1:4.70 (95% confidence interval, 3.65-6.07) for lidocaine/bupivacaine. In the next experiment, 45 rats intrathecally received 20 mu L of saline, 2.13% bupivacaine (approximately 1.5 mg/kg), or 10% lidocaine (approximately 6.9 mg/kg), and were examined for persistent functional impairment and morphologic damage. Rats given lidocaine developed significantly more prolonged tail-flick latencies than those in other groups 4 days after injection and incurred more morphologic damage than those given saline or bupivacaine. In conclusion, although the doses of anesthetics administered were larger than those used clinically, the present results suggest that bupivacaine is less neurotoxic than lidocaine when administered intrathecally at equipotent concentrations in the rat model.