OBJECTIVE Morphine is an effective analgesic, but adverse effects limit its clinical use in higher doses. The non-opioid antitussive, dextromethorphan (DM), can potentiate the analgesic effect of morphine and decrease the dose of morphine in acute postoperative pain, but the underlying mechanism remains unclear. We previously observed that DM increases the serum concentration of morphine in rats. Therefore, we investigated the effects of drugs administered at the spinal level to exclude possible pharmacokinetic interactions. As DM has widespread binding sites in the central nervous system [such as N-methyl-D-aspartate (NMDA) receptors, sigma receptors and alpha(3)ss(4) nicotinic receptors], we investigated whether the potentiation of morphine antinociception by DM at the spinal level is related to NMDA receptors. METHODS We used MK-801 as a tool to block the NMDA channel first, and then studied the interaction between intrathecal (i.t.) morphine and DM. The tail-flick test was used to examine the antinociceptive effects of different combinations of morphine and other drugs in rats. RESULTS DM (2-20 microg) or MK-801 (5-15 microg) showed no significant antinociceptive effect by themselves. The antinociceptive effect of morphine (0.5 microg, i.t.) was significantly enhanced by DM and reached the maximal potentiation (43.7%-50.4%) at doses of 2 to 10 microg. Pretreatment with MK-801 (5 or 10 microg, i.t.) significantly potentiated morphine antinociception by 49.9% or 38.7%, respectively. When rats were pretreated with MK-801, DM could not further enhance morphine antinociception (45.7% vs 50.5% and 43.3%). CONCLUSIONS Our results suggest that spinal NMDA receptors play an important role in the effect of DM to potentiate morphine antinociception.