The author reports the use of vertebral artery ligation, unilateral and bilateral, for the treatment of large vertebral-basilar aneurysms in 14 patients with one delayed death. Extracranial ligation was carried out unilaterally with a Selverstone clamp in three patients. In two, where the aneurysm filled only from one vertebral artery, there was extensive thrombosis within the sac and dramatic clinical improvement after decompression. Extracranial ligation was done bilaterally in three patients, temporarily in two. A 14-year-old boy is well after 5 years but the bilateral vertebrobasilar aneurysm did not undergo extensive thrombosis until both vertebral arteries were occluded at their intracranial entrance above collateral flow. In two others, the clamp had to be reopened on the second artery. In one patient, death from delayed thrombosis of a huge aneurysm and pontine infarction might have been prevented with anticoagulants. In the other, the aneurysm ruptured again fatally 18 months later. Unilateral intracranial occlusion of a vertebral artery was done in eight cases, with no morbidity and complete or nearly complete thrombosis in all but one aneurysm. Seven patients had excellent or good results while one showed little recovery from an existing medullary syndrome. Occlusion of the basilar artery was done in seven cases. In five it was used deliberately as the only treatment, but in two it was forced when an aneurysm burst during dissection. Only two of the patients in the first group and one of the second group have made complete recoveries. The results of vertebral artery occlusion are encouraging and the technique deserves further consideration. Extensive collateral circulation enhances the safety of cervical vertebral artery occlusion but can be of a degree to make the occlusion ineffective. For intracranial occlusion knowledge of the size and distribution of each vertebral artery is essential. Occlusion of the basilar artery is dangerous, although it seems to be effective in producing extensive thrombosis in the aneurysm. It should probably be done under anesthesia only when the artery fills spontaneously from the carotid circulation. Otherwise, even when reasonable posterior communicating arteries are demonstrated, it is best to test occlusion under local anesthesia.