No definite therapy for human stroke is currently available. Experimental stroke therapy so far has been based on the assumption that one, or at most, a few factors control the ischemic cell damage. There is a finite possibility that this assumption is wrong. Among the many factors participating in progression of neuronal cell loss, there may be none which are dominant. They may act instead as a network of minor causes, with intricate relationships among the various mediators and without clear sequence or directionality. The importance of considering this possibility is that the conventional search for a dominant mediator of ischemic cell death in stroke may be incompatible with the essential nature of the problem. As a new and alternative approach, we have started to investigate a state in nature, in which the tendency for cells exposed to cerebral ischemia to activate self-destructive processes, may be blunted, i.e., hibernation. Thirteen-lined ground squirrels were used as the hibernating species. Hibernation under laboratory conditions in this species was characterized by remarkable adaptive changes, including severe hypothermia, rapid onset bradycardia and hypotension, all of which are rapidly reversible. During deep hibernation, cerebral blood flow (CBF) levels were below the ischemic threshold, but no neuropathologic damage could be detected. The tendency to activate self-destructive processes in hibernation under conditions of low CBF is markedly reduced. Possible mechanisms of protection include reversible changes in white blood cells and platelets, changes in cell membrane fragility and maintenance of ion gradient homeostasis. Furthermore, we have preliminary evidence that humoral factors of such regulatory factors may enable us to prevent the breakdown of homeostatic mechanisms during cerebral ischemia in other species.