1. The reactions of nitric oxide with superoxide can lead to neurotoxicity through formation of peroxynitrite, and not by NO. alone, at least under our conditions. 2. Transfer of NO+ groups to thiol(s) on the NMDA receptor can lead to neuroprotection by inhibiting Ca2+ influx. These findings suggest that cell function can be controlled by, or through, protein S-nitrosylation, and raise the possibility that the NO group may initiate signal transduction in or at the plasma membrane. 3. The local redox milieu of a biological system is of critical importance in understanding NO actions as disparate chemical pathways involving distinct redox related congeners of NO may trigger neurotoxic or neuroprotective pathways. These claims are highlighted in the CNS by the recent finding that tissue concentrations of cysteine approach 700 microM in settings of cerebral ischemia (Slivka and Cohen, 1993); these levels of thiol would be expected to influence the redox state of the NO group. 4. Finally, our findings suggest novel therapeutic strategies. For example, downregulation of NMDA receptor activity via S-nitrosylation with NO+ donors could be implemented in the treatment of focal ischemia, AIDS dementia, and other neurological disorders associated, at least in part, with excessive activation of NMDA receptors.