A clinical analysis of vasospasm as seen angiographically after the onset of subarachnoid hemorrhage was carried out in 22 out of 4 cases of ruptured intracranial aneurysms experienced in our department from January, 1978 to May, 1979. The following results were obtained. The incidence of vasospasm was 50%. 2) Angiographic filling abnormalities(vasospasm) were morphologically classified into 4 types:tapering, diffuse, segmental and nodular types. 3) No significant difference was found in the location of the arteries with vasospasm in relation to the site of the ruptured aneurysm. In this study, the most frequent occurance of spasm was noted in cases of internal carotid aneurysm(69.2%). 4) Discrepancy in the incidence of angiographic spasm between the group with only one subarachnoid hemorrhage(41.9% out of 31 cases) and the other group with more than one hemorrhage(50% out of 44 cases) was demonstrated. While vasospasm occurred frequently within 9 days after the last hemorrhage in cases with more than two hemorrhage, the appearance of vasospasm was frequently seen angiographically between 10 to 17 days after hemorrhage in cases with only one hemorrhage. 5) There was no relationship between the age and the incidence of spasm, and most cases with spasm showed bloody and xanthochromic C.S.F. 6) Neurological findings were as follows:mental disturbance 12 cases, cranial nerve palsy 3 cases, paresis 3 cases, seizure 2 cases and visual disturbance 2 cases. These neurologic deficits may not be related to the spasm. 7) Follow up carotid angiography was performed in 9 cases of those treated by direct approach. Among 5 cases with vasospasm in preoperative angiogram, the spasm persited in 2 cases and was not no longer visualized in 3 cases. On the other hand, in 2 cases without preoperative spasm, postoperative spasm was detected. 8) Of 23 cases of ruptured aneurysms treated surgically(including 3 cases of carotid ligation), 2 out of 15 cases with spasm expired and all 8 cases without spasm were cured. 9) The diffuse type(using above classification) had the worst postoperative prognosis:i.e., among 9 cases with that type of spasm, 2 cases expired and one case resulted in a mild neurologic deficit. Other types of spasm had no special relationship with the postoperative prognosis.