The efficacy of the calcium antagonists nifedipine and nimodipine was evaluated in a multiinjection canine model of severe chronic cerebral vasospasm. Each of 21 adult mongrel dogs (15 to 20 kg) was assigned to one of four experimental groups. All animals received a total of 15 ml of fresh, unheparinized arterial blood via three cisterna magna injections. Selective left vertebral arteriography was performed 1 week before and exactly 7 days after the simulated subarachnoid hemorrhage. Treatment with calcium antagonists was initiated at 24 hours and continued for 7 consecutive days. Comparisons were based on the percentage of reduction in basilar artery diameter. On a milligram/kilogram basis, we used the manufacturer's maximal recommended daily dosages of nifedipine, whereas the dosages of nimodipine used were 3 to 7 times greater than that reported to be therapeutically effective in reducing the incidence of severe deficits from vasospasm in patients with aneurysmal subarachnoid hemorrhage. In this laboratory trial, systemic treatment with calcium antagonists did not prevent or significantly reduce chronic arterial narrowing. In the control group, spasm reduced basilar artery diameter an average of 71% +/- 11 (SD). In the group treated with nifedipine, 10 mg orally every 6 hours, vessel diameter was reduced an average of 58% +/- 13%. In those animals treated with nimodipine, 20 or 40 mg orally every 4 hours, the basilar artery diameter was decreased an average of 59% +/- 8 and 56% +/- 7, respectively. A preliminary trial using direct intracisternal administration of these agents demonstrated increased vessel diameter within 20 minutes in four of six animals, supporting other reports that this route of administration may prove more efficacious in the treatment of chronic cerebral vasospasm.