The purpose of this study was to provide further information on the usefulness of N-methyl-D-aspartate (NMDA) discrimination in rats as a behavioral model for NMDA receptor activation. The pharmacological specificity of the NMDA discriminative stimulus was examined in rats trained to discriminate 30 mg/kg, i.p. NMDA from saline using a 2-lever fixed-ratio (FR) 32 food reinforcement schedule. Pharmacologically diverse centrally-acting agents were examined for their ability to substitute for NMDA. Morphine did not substitute for NMDA; neither did the central stimulants, caffeine and (+)-amphetamine, which produced a maximum mean of only 16 and 35% NMDA-lever responding, respectively. Pentylenetetrazol and picrotoxin also did not substitute for NMDA. Compounds interacting with cholinergic neurotransmission including nicotine, physostigmine, arecoline and mecamylamine, produced at best, only intermediate levels of NMDA-lever responding (32-61 %), with the highest levels of NMDA-lever responding generally occurring at doses that also reduced rates of responding. These results suggest that the discriminative stimulus properties of NMDA are dissimilar from those of a number of centrally-acting drugs. Combined with the results of studies indicating that the NMDA discriminative stimulus can be antagonized by competitive NMDA antagonists, these results provide further evidence that NMDA receptor activation is the basis of NMDA discrimination and that this model may be useful for studying site-selective NMDA agonists and antagonists.