The neuroprotective effects of hypothermia during cerebral ischaemia or asphyxia are well known. Although, in view of this, the possibility of a therapeutic role for hypothermia during or after resuscitation from such insults has been a long standing focus of research, early studies had limited and contradictory results. Clinically and experimentally severe perinatal asphyxial injury is associated with a latent phase after reperfusion, with initial recovery of cerebral energy metabolism but EEG suppression, followed by a secondary phase with seizures, cytotoxic edema, accumulation of cytotoxins, and cerebral energy failure from 6 to 15 h after birth. Recent studies have led to the hypothesis that changes in post-ischaemic cerebral temperature can critically modulate encephalopathic processes which are initiated during the primary phase of hypoxia-ischaemia, but which extend into the secondary phase of cerebral injury. This conceptual framework allows a better understanding of the 'pharmacological' parameters that determine effective hypothermic neuroprotection, including the timing of initiation of cooling, its duration and the depth of cooling attained. Moderate cerebral hypothermia initiated in the latent phase, between one and as late as 6 hours after reperfusion, and continued for a sufficient duration in relation to the severity of the cerebral injury, has been associated with potent, long-lasting neuroprotection in both adult and perinatal species. These encouraging results must be balanced against the adverse systemic effects of hypothermia. Randomised clinical trials are in progress to establish the safety and efficacy of prolonged cerebral hypothermia.