Data from rodent antinociception models indicate that N-methyl-D-aspartate (NMDA) receptor antagonists do not produce antinociception alone or potentiate morphine antinociception, but do attenuate the development of morphine tolerance. This study examined the antinociceptive effects of the noncompetitive NMDA receptor antagonist dizocilpine, the competitive NMDA receptor antagonist (-)-6-phosphonomethyl-decahydroisoquinoline-3-carboxylic acid (LY235959), and the glycine-site antagonist (+)-(1-hydroxy-3-aminopyrrolidine-2-one) [(+)-HA-966], alone and in combination with morphine in a squirrel monkey titration procedure. In this procedure, shock (delivered to the tail) increased in intensity every 15 s from 0.01 to 2.0 mA in 30 increments. Five lever presses during any given 15-s shock period produced a 15-s shock-free period after which shock resumed at the next lower intensity. Morphine (0.3-3.0 mg/kg i.m.) dose-dependently increased the intensity below which monkeys maintained shock 50% of the time (median shock level; MSL). In contrast, dizocilpine (0.003-0.1 mg/kg i.m.) produced only modest increases in MSL in some monkeys (three of five) at the highest dose tested. Neither LY235959 (0.1-1.0 mg/kg i.m.) or (+)-HA-966 (10-56 mg/kg i.m.) increased MSL in any monkey tested. Dizocilpine, LY235959, and (+)-HA-966, when administered in combination with doses of morphine (1.0 mg/kg, 1.7 mg/kg) that either produced no antinociception or produced very little antinociception, were all found to dose-dependently potentiate the antinociceptive effect of morphine. Importantly, although these NMDA antagonists in combination with morphine produced marked increases in MSL, these combinations did not alter response rate, demonstrating that the potentiation was not due to nonspecific motor effects.