Thirty-one monkeys were randomly divided into three groups to undergo baseline cerebral angiography followed by induction of subarachnoid hemorrhage by placement of autologous blood clot along the right-sided arteries of the anterior circle of Willis (Day 0). The monkeys were then given drug vehicle or one of two endothelin (ET) antagonists, BQ-123 (6 mg/kg/day) or bosentan (5 mg/kg/day) intracisternally. The BQ-123 was administered by continuous infusion from a subcutaneous pump and the bosentan was given by twice-daily injections into an Ommaya reservoir in the subcutaneous space with a catheter along the right middle cerebral artery (MCA). Seven days later (Day 7), angiography was repeated and the animals were killed. Comparison of arterial diameters shown on angiograms between Day 0 and Day 7 groups given placebo and bosentan showed significant reductions in the diameters of the right intradural internal carotid (28% +/- 6% and 30% +/- 6%, respectively, paired t-test, p < 0.05), anterior cerebral artery (29% +/- 8% and 32% +/- 6% respectively +/- 6%, respectively) and MCA (34% +/- 6% and 46% +/- 4%, respectively). Animals injected with BQ-123 had significant narrowing of the right extradural internal carotid artery (7% +/- 6%) and the basilar artery (11% +/- 3%), but not of the right MCA. Comparison of arterial diameters between groups at Day 7 showed significant variance in the right extradural internal carotid, both intradural internal carotid, right middle cerebral, and left anterior cerebral arteries; the animals injected with BQ-123 developed significantly less arterial narrowing these those receiving bosentan and placebo. Bosentan was not detected in the cerebrospinal fluid aspirated from the cisterna magna on Day 7, whereas BQ-123 was detected in two animals. We can infer from these results that BQ-123 prevents vasospasm following subarachnoid hemorrhage in monkeys, that further investigations of ET antagonists are warranted, and that ET may be an important pathophysiological mediator of vasospasm. The lack of efficacy of bosentan may be related to inadequate cerebrospinal fluid levels obtained by administration twice-daily through an Ommaya reservoir.