Over a 2-year period the authors have studied the effects of preoperative subselective embolization of meningiomas. Thirty-six consecutive patients shown by computerized tomography (CT) to have a meningioma underwent selective internal and external carotid artery angiography, and any significant external carotid artery feeders were embolized (27 cases). It was found that CT and dynamic radioisotope scan findings were unable to predict the degree of vascularity of the tumor or its suitability for embolization. Furthermore, these tests, repeated after embolization, were unreliable in detecting either the extent of necrosis or reduction in blood flow. The effects of embolization upon surgery were recorded, and the excised tumor specimen examined for evidence of thrombosis and infarction. Subselective embolization was determined to be a simple, safe, and effective method of producing tumor necrosis and intraoperative hemostasis in selected patients.