Between 1980 and 1987, 1622 patients with angiographically verified ruptured cerebral aneurysms were admitted within 7 days after subarachnoid hemorrhage. A retrospective analysis evaluated both the timing of surgery in operative patients and the status of nonsurgical patients. The patients' clinical grade according to the Hunt and Hess classification was assessed at admission, and a comparative analysis of outcome was carried out for each grade in relation to time of surgery: those operated on from Day 0 to 3 and those undergoing surgery on Day 4 or later. Among nonsurgical cases, fatal rebleeding occurred in 105 cases and fatal vasospasm in 69 cases. These nonsurgical cases were divided into one of two groups, either an early- or a late-management group, and the outcome of each group was analyzed by clinical grade. The mortality rates in the early-surgery groups were higher than in the late-surgery groups, especially in Grade V, in which the rate was significantly different. However, with the 174 nonsurgical patients included in these management results, marked differences in mortality rates disappeared except in Grade V, which failed to show statistical significance. A higher rate of good recovery among Grade III patients receiving early surgery shifted significantly in the early-management group. The results suggest that the timing of surgery in clinical Grade I or II patients is not a major factor; however, early surgery appears to be beneficial in Grade III and IV patients. The incidence of rebleeding in the early- and late-management groups was 2.7% and 9.5%, respectively.