Hypothermia has long been employed for therapeutic purposes but its use has been limited because of the potential life-threatening side-effects. In late eighties a neuroprotective effect of bold hypothermia was demonstrated and this implied that the method could be used more safely. It has then been shown in a lot of animal experiments that post-ischaemic mild hypothermia significantly limits damage to the brain caused by cardiac arrest. Similar results have been obtained for local brain ischaemia and for experimental head trauma. Molecular basis for this neuroprotection with mild hypothermia has been found to be complex, involving attenuation of the excitotoxic effects of glutamate and diminishing the synthesis of free radicals. In the last decade some clinical series and multicenter randomised trials have shown that mild hypothermia is safe and effective in global brain ischaemia due to cardiac arrest. Clinical data suggest also its effectiveness in ischaemic stroke though no multicenter randomised study has been published to date. At present there are conflicting results of clinical trials concerning brain injuries. Although some authors have reported up to 38% improvement in the outcome, a recently published multicenter randomised trial has failed to demonstrate any practical benefit of mild hypothermia after acute brain injury. There is however virtually no data in the literature on mild hypothermia in spinal cord injury.