The pharmacologic specificity and anatomic distribution of [3H]dextrorphan recognition sites in the rat brain was characterized by quantitative autoradiography. Equilibrium saturation analysis indicated that [3H]dextrorphan labeled a single population of high affinity binding sites. These sites are heterogeneously distributed throughout rat forebrain with the following order of binding densities: hippocampal formation > cerebral cortex > thalamic nuclei > striatum. The association rate of [3H]dextrorphan with its binding site in area stratum radiatum of CA1 is accelerated by the addition of glycine and glutamate. [3H]Dextrorphan binding is, however, relatively insensitive to glycine and glutamate under equilibrium conditions, despite extensive prewashing procedures to deplete endogenous levels of these substances. The competitive N-methyl-D-aspartate (NMDA) antagonist D(-)-2-amino-5-phosphonopentanoic acid (D-AP5) and the glycine site antagonist 7-chlorokynurenic acid completely inhibit specific [3H]dextrorphan binding. D-AP5 suppresses [3H]dextrorphan binding in a regionally distinctive manner; a population of binding sites is weakly inhibited by D-AP5 in the lateral thalamic regions, whereas D-AP5 potently inhibits [3H]dextrorphan binding in the cerebral cortex. The rank order of potencies of an array of noncompetitive antagonists to inhibit [3H]dextrorphan binding unambiguously displays the pharmacologic profile of the noncompetitive antagonist domain of the NMDA receptor-channel complex. Furthermore, the distribution of [3H]dextrorphan binding sites in slide-mounted tissue appears qualitatively similar to the distribution of NMDA receptors previously reported using NMDA-displacement of [3H]glutamate, [3H](+)-5-methyl-10,11-dihydro-5H-dibenzo[a,d]cyclohepten-5,10-imi ne (MK-801) and [3H]1-[1-(2-thienyl)cyclohexyl]-piperidine (TCP) in most brain areas examined except the cerebellum. The molecular layer of the cerebellum displays a particularly high density of [3H]dextrorphan binding sites. The regional distribution of [3H]dextrorphan binding sites in rat brain does not correspond to the reported distributions of [3H]dextromethorphan or sigma binding sites.