Of 43 cavernous sinus aneurysms diagnosed over 6 1/2 years, 23 fulfilled indications for treatment; of these 19 were treated, eight surgically and 11 with interventional radiological techniques. Six small and two giant aneurysms were treated surgically: four were clipped, two were repaired primarily, and two were trapped with placement of a saphenous-vein bypass graft. Seven large and four giant aneurysms were treated with interventional radiological techniques: in five cases the proximal internal carotid artery (ICA) was sacrificed; one aneurysm was trapped with detachable balloons; and five were embolized with preservation of the ICA lumen. The mean follow-up period was 25 months. At follow-up examination, three patients in the surgical group were asymptomatic, two had improved, and three had worsened. Three of these patients had asymptomatic infarctions apparent on computerized tomography (CT) scans. At follow-up examination, four radiologically treated patients were asymptomatic, five had improved, two were unchanged, and none had worsened. One patient had asymptomatic and one minimally symptomatic infarction apparent on CT scans; both lesions were embolic foci after aneurysm embolization with preservation of the ICA. It is concluded that treatment risk depends more on the adequacy of collateral circulation than on the size of the aneurysm. A multidisciplinary treatment protocol for these aneurysms is described, dividing patients into high-, moderate-, and low-risk groups based on pretreatment evaluation of the risk of temporary or permanent ICA occlusion using a clinical balloon test occlusion coupled with an ICA-occluded stable xenon/CT cerebral blood flow study. Radiological techniques are suggested for most low-risk patients, while direct surgical techniques are proposed for most moderate- and high-risk patients.