Surgical or nonsurgical treatment was randomly selected for 1,225 patients with occlusive cerebrovascular disease. Statistically significant differences in cumulative survival rate at 42 months were as follows: (1) Surgical treatment appeared more beneficial for unilateral carotid artery stenosis in patients with transient attacks or a mild-to-moderate neurological deficit. (2) Nonsurgical treatment produced better results for unilateral carotid artery stenosis in patients with a moderateto-severe neurological deficit. (3) Nonsurgical treatment appeared more beneficial for combined unilateral carotid artery stenosis and contralateral carotid artery occlusion if patients had a moderate-to-severe neurological deficit. (4) Nonsurgical treatment appeared more beneficial for patients with completed strokes who had a marked and persistent neurological deficit.