Beginning with electrophysiological evidence for two populations of receptors for N-methyl-D-aspartate (NMDA) which did or did not respond to the agonist quinolinic acid, evidence has grown for such subdivision. 2. Data from binding studies is consistent with differences between three NMDA receptors in the striatum, thalamus and cerebellum with respect to their preferences for agonist or antagonist binding and the modulation of binding by dizocilpine, cations and polyamines. 3. The recent isolation and sequencing of several different molecular species of NMDA receptor supports the view that at least two pharmacologically distinct sites exist, with the cerebellar receptor being unique in the brain.