The authors have developed a method to induce chronic cerebral vasospasm after subarachnoid hemorrhage (SAH) in monkeys. With microsurgical techniques, 33 monkeys had a frontotemporal craniectomy and unilateral opening of the subarachnoid cisterns. Cerebrospinal fluid was drained and a fresh hematoma, obtained from an average of 7 ml of autologous blood, was carefully placed against the major arteries of the anterior circulation on one side. The 30 monkeys studied for 7 to 14 days after the SAH were allocated randomly to two treatment groups of 15: one group received placebo and the other nimodipine, 1 mg/kg every 8 hours. Indices monitored before and after SAH included neurological status, cerebral blood flow, computerized tomography, and angiographic vessel caliber. In the placebo group, delayed ischemic neurological deficit developed in one monkey 4 days after clot placement and was present at sacrifice on Day 14. No such deficit occurred in the nimodipine group. The effect of nimodipine on vessel caliber at this dosage was equivocal. Significant vasospasm (31% to 100% reduction in vessel caliber) developed in 87% (26 of 30) of the animals. Overall, vasospasm was slightly more common in the placebo group: in this group, on Days 7 and 14, the incidence of vasospasm was significantly higher (p less than 0.05) than in the nimodipine group. However, the average percentage reduction in vessel caliber of the maximally constricted vessel in each monkey was not significantly different between the two groups.