Among 250 patients consecutively admitted in our center with a ruptured intracranial aneurysm, 66 patients (24.4%) were initially classified in Hunt and Hess clinical grade IV (37 cases) or grade V (29 cases). All patients were studied as following: --The severity of subarachnoid haemorrhage was evaluated on the pre-operative C.T. scan using Fisher's criteria. Quantification of the intracerebral haematoma or the intraventricular associated haemorrhage or a subdural haematoma was estimated as well. --The arterial diameter, aneurysm size and location, and the eventual presence of intra-arterial embolus were noted on the pre-operative angiography. Aneurysm location was: anterior artery 27 cases, internal carotid artery 9 cases, middle cerebral artery 27 cases, posterior cerebral artery 2 cases, and 1 case on the basilar artery. In this series, a poor clinical condition on admission was related to the direct effect of the initial haemorrhage in 84.9% of the patients. Early seizures (7.6%) acute hydrocephalus (1.5%), multiple emboli (3%) and apparently early diffuse vasospasm (1.5%) were the other documented causes explaining the initial poor clinical condition. One case remained completely unexplained. Sixteen patients admitted with bilateral fixed dilated pupils or a major intracerebral haematoma from a ruptured anterior artery aneurysm were not operated on and subsequently died. Operative treatment (aneurysm clipping in all cases, and haematoma evacuation on demand) was performed in the remaining 50 cases within 12 hours after their admission. Thus, these patients underwent surgery on Day 0 in 31 cases, on Day 1 in 11 cases, and on Day 2 in 8 cases. A post-operative C.T. scan was performed in 46 cases. Post-operative angiographic control was only performed in 29 cases. Changes on the post-operative C.T. scan or the angiographic control were strictly compared to the neuroradiological information previously available. Final outcome was assessed at least two months after the onset. According to the Glasgow Outcome Score, the results were: good recovery 12 patients (18.2%); moderate disability 1 patient; severe disability 4 patients (6%); vegetative state 4 patients; death 45 patients (68.2%). Excluding the patients admitted in poor clinical grade but presenting with early seizures or minor initial haemorrhage, the mortality rate was 74.2%. According to the initial clinical grade, the initial C.T. scan findings, the eventual post-operative angiographic presence of an arterial thrombosis or vasospasm, it was obvious that the final bad outcome was mainly related to the severity of the initial haemorrhage. However, in 7 patients, post-operative disability or death can be explained by other complications, principally an arterial thrombosis.(ABSTRACT TRUNCATED AT 250 WORDS)