We investigated the clinical features of 54 elderly patients aged over 80 years with ruptured intracranial aneurysms admitted to our hospital between May 1968 and December 2001. The changes in patient number, age, male/female ratio, site of cerebral aneurysm and preoperative Hunt & Kosnik grade were studied. The patients were divided into 2 groups (surgically treated group and conservatively treated group). Outcome was evaluated according to the Glasgow Outcome Scale, and the causes of poor results were analyzed.The number of patients calculated for 5-year periods is rising steadily, and reached 25 patients for the last 5 years (1997-2001), accounting for 46% of all. The mean age of the patients was 82.4 ± 2.5 (range 80-92 years), and 44 patients (81%) were under 84 years. The male/female ratio was 1:5. The location of 50 ruptured cerebral aneurysms (4 cases were excluded because cerebral angiography was not performed) were the internal carotid artery in 24 cases (48%), the middle cerebral artery in 12 cases (24%), the anterior communicating artery in 11 cases (22%), the anterior cerebral artery in 2 cases (4%), and the basilar artery in 1 case (2%). Their preoperative neurological grades, according to the classification of Hunt & Kosnik, were II in 11 cases (20%), III in 17 cases (31%), IV in 14 cases (26%) and V in 12 cases (22%).Of 54 cases, 26 were treated surgically: neck clipping of cerebral aneurysms was performed in 21 cases, wrapping of cerebral aneurysms in 2 cases, transarterial embolization of cerebral aneurysms in 3 cases and the remaining 28 cases were treated conservatively. The clinical outcome of the surgically treated group was GR in 4 cases (15%), MD in 5 cases (19%), SD in 6 cases (23%), VS in 2 cases (8%) and D in 9 cases (35%). The preoperative Hunt & Kosnik grades of the 4 GR cases were II in 1 case, III in 2 cases, and IV in 1 case. Cerebral infarction caused by cerebral vasospasm was revealed in 10 (38%) of 26 surgically treated cases. The preoperative Hunt & Kosnik grades of these 10 cases were II in 4 cases and III in 6 cases. The clinical outcome was MD in 1 case, SD in 1 case, VS in 1 case and D in 7 cases. Cerebral infarction caused by cerebral vasospasm during the last 5 years (1997-2001) was revealed in 3 (23%) of 13 surgically treated cases, although it was 7 (54%) of 13 cases before 1996. Twenty-two of 28 (79%) conservatively treated cases were serious (Hunt & Kosnik grade IV, V). The clinical outcome was GR in 1 case, SD in 2 cases, VS in 3 cases and D in 22 cases. Causes of death were primary brain damage in 14 cases, re-rupture in 4 cases, cerebral vasospasm in 3 cases and pneumonia in 1 case. It is necessary for determination of surgical indication to evaluate primary brain damage more strictly. And it is indispensable for improvement of operative outcome to take measures for cerebral vasospasm and general complications.