1. Quinolinic acid may be an important endogenous excitotoxin, but its concentrations in brain are low. We have therefore attempted to determine whether its neurotoxicity can be increased by the simultaneous presence of free radicals. 2. Quinolinic acid was injected into the hippocampus of anaesthetized rats at doses of 40 and 80 nmols which produced little neuronal loss, and 120 nmols which produced over 90% neuronal loss. 3. A mixture of xanthine and xanthine oxidase, a known source of free radical reactive oxygen species, also generated little damage alone, but killed over 80% of CA1 neurons when combined with 80 nmols of quinolinic acid. Similarly, the nitric oxide donor S-nitroso-N-acetylpenicillamine (SNAP) potentiated the damage produced by quinolinic acid. 4. The glutamate antagonist 5,7-dichlorokynurenic acid prevented the damage produced by 120 nmols of quinolinic acid, but not that produced by quinolinic acid plus xanthine/xanthine oxidase, indicating that damage was not simply the result of free radical enhancement of NMDA receptor activation. 5. Three chemically dissimilar antagonists at adenosine A(2A) receptors prevented the damage caused by quinolinic acid and xanthine/xanthine oxidase or by quinolinic acid plus SNAP. 6. It is concluded that reactive oxygen species can potentiate the neurotoxicity of quinolinic acid. The site of interaction is probably distal to the NMDA receptor. Blockade of adenosine A(2A) receptors can protect against this combined damage, suggesting potential value in the prevention of brain damage.