Hypothermia was first applied therapeutically as a local anesthetic and later was used to achieve organ protection during procedures necessitating circulatory interruption. Profound whole-body hypothermia, typically carried out in conjunction with extracorporeal bypass, has long been employed during cardiac and neurosurgical operative procedures. More recently, studies in small-animal experimental models of cerebral ischemia have provided persuasive evidence that even small decreases in brain temperature confer striking protection against ischemic neuronal injury. By contrast, small elevations of brain temperature during ischemia accelerate and extend pathologic changes in the brain and promote early disruption of the blood-brain barrier. Hypothermia retards the rate of high-energy phosphate depletion during ischemia and promotes postischemic metabolic recovery. More importantly, mild intraischemic hypothermia markedly attenuates the release of glutamate into the brain's extracellular space and significantly diminishes the release of dopamine. Similarly, the inhibition of calcium-calmodulin-dependent protein kinase II triggered by normothermic ischemia is prevented by hypothermia, as is the ischemia-induced translocation and inhibition of the key regulatory enzyme protein kinase C. Hypothermia also appears to facilitate the resynthesis of ubiquitin following ischemia. Studies of potential clinical importance have shown that moderate hypothermia is capable of attenuating ischemic damage even if instituted early in the postischemic period. In the setting of focal cerebral ischemia, moderate brain hypothermia reduces the infarct size (particularly in the setting of reversible middle cerebral artery occlusion); conversely, hyperthermia markedly increases the infarct volume. These studies underscore the importance of monitoring and regulating the brain temperature during experimental studies of cerebral ischemia to insure a consistent pathologic outcome and to avoid the false attribution of "pharmacoprotection" to drugs that reduce the body temperature. The measurement of brain temperature is now practicable in neurosurgical patients requiring invasive monitoring, and human studies have shown that cortical and cerebroventricular temperatures may exceed systemic temperatures. Mild to moderate decreases in brain temperature are neuroprotective in cerebral ischemia, while mild elevations of brain temperature are markedly deleterious in the setting of ischemia or injury. It is anticipated that controlled clinical trials of therapeutic brain temperature modulation will be undertaken over the next several years.