The results of surgical evacuation of spontaneous intracerebral hematomas are disappointing. This is largely because experimental studies have now confirmed that the brain surrounding an intracerebral hematoma develops profound and extensive ischemia. The volume of this ischemic brain may exceed the volume of the hemorrhage several times. This has been demonstrated experimentally using 14C-iodoantipyrene autoradiography in various modifications of the intracerebral hemorrhage model. These models have demonstrated that the pathophysiology of the ischemia is partly due to direct mechanical compression. There is also a component of the ischemic process induced by vasoconstrictor substances in blood. The diffuse uncontained type of hemorrhage (subarachnoid or intraventricular) causes a global reduction in cerebral perfusion pressure. The focal ischemic event is initiated at the time of hemorrhage and is largely irreversible. The experimental evidence to date indicates that neuroprotective agents (calcium channel blockers and N-methyl-D-aspartate receptor antagonists) reduce ischemic brain damage. Similarly, in immunosuppressed animals the amount of brain edema that follows the initial ischemic insult was reduced. These studies indicate that pharmacological neuroprotective strategies can minimize the brain damage that follows intracerebral hemorrhage. Early removal of the mass lesion may play a role, but it is unlikely to reverse the ischemic process if it is the only treatment offered.