Quantitative concentration-toxicity relationships were determined for the injury of cultured murine cortical neurons by several excitatory amino acid (EAA) agonists. All tested agonists produced concentration-dependent neuronal injury at concentrations between 1 and 1000 microM. With 5 min exposure, glutamate, aspartate, N-methyl-D-aspartate (NMDA), L-homocysteate (HCA), and quisqualate all had similar potencies, destroying half of the neuronal population (LD50) at concentrations of 50-200 microM, and similar efficacies, with 88-92% neuronal loss produced by exposure to high agonist concentrations. Quinolinate and kainate were substantially weaker toxins, producing only 20-30% neuronal loss after 5 min exposure to 3 mM concentrations; with prolonged (24 hr) exposure, 85-95% neuronal loss could be attained. The comparative EAA vulnerability of a specific cortical neuronal subpopulation containing high concentrations of the enzyme, reduced nicotinamide adenine dinucleotide phosphate diaphorase (NADPH-d), was also examined. Glutamate had no differential toxicity on these cells, damaging them at all concentrations in proportion to the general population; however, other, more selective, agonists produced strikingly differential injuries. These NADPH-d-containing [NADPH-d(+)]neurons were selectively resistant to damage by low concentrations of the NMDA agonists quinolinate, HCA, aspartate, or NMDA itself. By contrast, NADPH-d(+)neurons were selectively destroyed by concentrations of quisqualate or kainate too low to produce much general neuronal injury. The differential susceptibility of these neurons was not absolute, as high concentrations of all tested agonists produced nonselective neuronal injury. In light of recent evidence that forebrain NADPH-d(+)neurons are selectively spared in Huntington's disease, the present study continues to support the hypothesis that neuronal loss in Huntington's disease might result from excessive NMDA-receptor stimulation by any selective NMDA agonist. Furthermore, the demonstration that the differential susceptibility of NADPH-d(+)neurons is agonist concentration-dependent, rather than absolute, could provide a basis for explaining some existing conflicting experimental data.