Although hypothermic circulatory arrest is used frequently in cardiac surgery, questions remain about the adequacy of brain protection provided and the duration of circulatory arrest allowable. Some surgeons operating on young children have utilized hypothermic low flow instead of complete circulatory arrest, in an attempt to minimize the risk of cerebral injury, although this has never been demonstrated beneficial. Similarly, some surgeons operating on the elderly for pulmonary thrombosis have used intermittent reperfusion in an attempt to safely extend the period of arrest, although there is no evidence to indicate this is useful. In actuality, experimental data at normothermic conditions suggests that repeated reperfusion or low flow could be injurious and our preliminary experiments indicate that low flow causes more injury than arrest for the same period. A major drawback to animal studies on brain protection is the difficulty in examining intellectual activities, such as memory, which are clearly important to patients. Therefore we have elected to use a morphological marker of neural damage in the hippocampal neurons which are known to be most sensitive to ischemia to determine if repeated re-perfusions or low flow are beneficial or detrimental to experimental animals. Extensive literature on normothermic cerebral ischemia suggests that glucose depletion may increase the brain tolerance to circulatory arrest. We propose to combine hypoglycemia and hypothermia in an attempt to extend the allowable period of circulatory arrest. Recent animal experiments suggest that neuroprotective agents exist which may ameliorate the neuronal injury associated with global ischemia. We propose to pretreat animals undergoing hypothermia circulatory arrest with some of the most promising of these agents to determine whether they will have a neuroprotective effect in our model.