The role of the neural messenger nitric oxide (NO) in cerebral ischaemia has been investigated extensively in the past decade. NO may play either a protective or destructive role in ischaemia and the literature is plagued with contradictory findings. Working with NO presents many unique difficulties and here we review the potential artifacts that may have contributed to discrepancies and cause future problems for the unwary investigator. Recent evidence challenges the idea that NO from neurones builds up to levels (micromolar) sufficient to directly elicit cell death during the post-ischaemic period. Concomitantly, the case is strengthened for a role of NO in delayed death mediated post-ischaemia by the inducible NO synthase. Mechanistically it seems unlikely that NO is released in high enough quantities to inhibit respiration in vivo; the formation of reactive nitrogen species, such as peroxynitrite, represents the more likely pathway to cell death. The protective and restorative properties of NO have become of increasing interest. NO from endothelial cells may, via stimulating cGMP production, protect the ischaemic brain by acutely augmenting blood flow, and by helping to form new blood vessels in the longer term (angiogenesis). Elevated cGMP production may also stop cells dying by inhibiting apoptosis and help repair damage by stimulating neurogenesis. In addition NO may act as a direct antioxidant and participate in the triggering of protective gene expression programmes that underlie cerebral ischaemic preconditioning. Better understanding of the molecular mechanisms by which NO is protective may ultimately identify new potential therapeutic targets.