The first 48 hours after aneurysmal subarachnoid hemorrhage are critical in determining final outcome. However, most patients who die during this initial period are not included in hospital-based studies. We investigated the occurrence of subarachnoid hemorrhage in a population-based study to evaluate possible predictors of poor outcome. All patients diagnosed with aneurysmal subarachnoid hemorrhage between 1955 and 1984 were selected for analysis of mortality in the first 30 days using the medical record-linkage system employed for epidemiological studies in Rochester, Minnesota. One hundred and thirty-six patients were identified. The mean age of these 99 women and 37 men was 55 years. Rates for survival to 48 hours were 32% for the 19 patients with posterior circulation aneurysms, 77% for the 87 patients with anterior circulation aneurysms, and 70% for the 30 patients with a presumed aneurysm (p < 0.0001). Rates for survival to 30 days were 11%, 57%, and 53%, respectively, in these three patient groups (p < 0.0001). Clinical grade on admission to the hospital, the main variable predictive of death within 48 hours, was significantly worse in patients with posterior circulation aneurysms than in others (p < 0.0001). The prognosis of ruptured posterior circulation aneurysms is poor. The high early mortality explains why posterior circulation aneurysms are uncommon in most clinical series of patients with subarachnoid hemorrhage. The management of incidentally discovered intact posterior circulation aneurysms may be influenced by these findings.