During cerebral ischaemia, glutamate is released in supraphysiological amounts and is toxic to brain tissue. This excitotoxicity is mediated by several glutamate receptor subtypes, including the ionotropic N-methyl-D-aspartate (NMDA) and alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) receptors. Clinical trials of drugs that block the NMDA receptor in acute ischaemic stroke have been disappointing. No improvement in clinical outcome of stroke has been seen with competitive NMDA antagonists (selfotel) and non-competitive NMDA antagonists (dextrorphan, GV150526, aptiganel and eliprodil). The AMPA receptor differs in important ways from the NMDA receptor. It is the principal mediator of fast excitatory neurotransmission. This ligand-gated cation channel is primarily permeable to sodium rather than calcium. It is found in grey and white matter. It is expressed by oligodendrocytes. This distribution may provide neuroprotection for both grey and white matter. In a variety of animal models, reduction in infarct volume with AMPA blockade has been demonstrated. AMPA antagonists also show benefit in spinal cord ischaemia and trauma. The clinical development of safe and effective AMPA blockers has been hampered by poor water solubility and associated renal toxicity. A novel, highly water-soluble, competitive AMPA receptor antagonist, YM872 ([2,3-dioxo-7-(1H-imidazol-1-yl)-6-nitro-1,2,3,4-tetrahydroquinoxalin-1-yl]acetic acid monohydrate; Yamanouchi), has been identified. Phase I clinical trial data indicate that this agent can be safely administered in young and elderly subjects. Sedation and other CNS associated adverse events determine the ceiling dose and become more problematic with infusion times exceeding 24 h. Phase II studies of YM872 in acute ischaemic stroke are ongoing.