Decreased endothelium-derived relaxing factor, nitric oxide (NO), in the arterial wall has been hypothesized to be a potential cause of cerebral vasospasm following subarachnoid hemorrhage (SAH). The authors sought to determine whether intracarotid infusions of newly developed NO-donating compounds (NONOates) could reverse vasospasm or prevent the occurrence of cerebral vasospasm in a primate model of SAH. Twenty-one cynomolgus monkeys were studied in two experimental settings. In an acute infusion experiment, saline or NONOate was infused intracarotidly in four normal monkeys and in four monkeys after onset of SAH. During the infusions regional cerebral blood flow (rCBF) was measured in eight animals and CBF velocity in two. In a chronic infusion experiment, saline (four animals) or NONOate (diethylamine-NO [three animals] or proli-NO [six animals]) was infused intracarotidly in monkeys for 7 days after SAH. In acute infusion experiments, 3-minute intracarotid diethylamine-NO infusions reversed arteriographically confirmed vasospasm of the right middle cerebral artery (MCA) (as viewed on anteroposterior projection, the decrease in area was 8.4+/-4.3% in the treatment group compared with 35+/-12% in the control group; p < 0.004), increased rCBF by 31+/-1.9% (p < 0.002), and decreased the mean systolic CBF velocity in the right MCA. In a long-term infusion experiment, the area of the right MCA in control animals decreased by 63+/-5%. In animals undergoing a 7-day continuous glucantime-NO intracarotid infusion, the area of the right MCA decreased by 15+/-6.2%, and in animals undergoing a 7-day proli-NO infusion, the area of the right MCA decreased by 11+/-2.9% (p < 0.05). The mean arterial blood pressure decreased in the glucantime-NO group from 75+/-12 mm Hg (during saline infusion) to 57+/-10 mm Hg (during glucantime-NO infusion; p < 0.05), but it was unchanged in animals undergoing proli-NO infusion (76+/-12 mm Hg vs. 78+/-12 mm Hg). Results of these experiments show that cerebral vasospasm is both reversed and completely prevented by NO replacement. However, only the use of regional infusion of the NONOate with an extremely short half-life avoided a concomitant decrease in arterial blood pressure, which could produce cerebral ischemia in patients with impaired autoregulation of CBF after the rupture of an intracranial aneurysm.