BACKGROUND The rebleeding of a ruptured intracranial aneurysm in the acute stage has been thought to indicate a very poor prognosis. In our experience, the outcome of patients with extravasation of the contrast medium is worse than that of patients with rerupture under circumstances other than angiography. We demonstrated the poorer outcome of the patients with extravasation, compared to that of the patients developing rerupture under circumstances other than angiography, and examined the factors that contributed to the poorer outcome of the patients with extravasation. METHODS Among the 641 cases of ruptured intracranial aneurysm, we have handled over the past 10 years, 36 (5.6%) patients have rebled before surgery was performed. Thirteen patients rebled during angiography (Group I), and the remaining 23 patients rebled in other circumstances (Group II). We compared the outcome of both groups. RESULTS The rebleeding occurred within 6 hours of the initial rupture in 29 (80%) patients. The outcome of Group I was significantly poorer than that of Group II (p < 0.05). In Group I, SD was achieved in only 1 patient and the remaining 12 patients died, while in Group II, 8 patients were in GR or MD, 2 were in SD, and the remaining 13 patients died. CONCLUSIONS Because rebleeding during angiography most often occurs in the acute stage and because the outcome of patients with extravasation is very poor, we recommend that the performance of angiography be delayed at least between 3 and 6 hours after the initial rupture.