NMDA receptor antagonists have been shown to be anxiolytic in animal models of anxiety, although they have not been tested extensively. These compounds bind to several specific sites within the NMDA-receptor complex, including the NMDA site itself, the phencyclidine site, and the strychnine-insensitive glycine site. The purpose of the present study was to examine potential anxiolytic effects of site-selective NMDA receptor antagonists in the elevated plus-maze. Drug-naive albino mice were placed in the center of an elevated maze shaped like a plus sign. Two opposing arms were enclosed by high walls; the crossing arms were open. Following injection with drug or vehicle, the number of entries and time spent in each type of arm were measured during 5-min tests. Analysis of results showed that the benzodiazepine, diazepam, and the competitive NMDA receptor antagonist, NPC 17742 (2R,4R,5S 2-amino-4,5-(1,2-cyclohexyl)-7-phosphono-heptanoic acid), increased number of open arm entries and open arm time. N-Nitro-L-arginine methyl ester, a nitric oxide synthase inhibitor which may interfere with the transduction of NMDA receptor activation, also increased open arm entries and time; however, the magnitude of these increases was small. The phencyclidine-site NMDA receptor antagonist, phencyclidine, increased open arm entries, but failed to significantly increase open arm time. ACEA 1021 (5-nitro-6,7-dichloro-1,4-dihydro-2,3-quinoxalinedione), a putative glycine-site antagonist, had significant effects only on open arm entries at the highest dose tested. These results suggest that NMDA receptor antagonists show promise as potential anxiolytic agents, but that differences among antagonists acting at different cellular sites may be expected.