Thiopental (10 mg/kg iv and 10 mg/kg im), phenytoin (15 mg/kg iv), and placebo were given to three groups of 12 anesthetized rabbits each before a pneumatic tourniquet was placed around the neck and inflated to 250 mm Hg for 20 min. No pulsations were noted in the retinal or auricular arteries, and the EEG was flattened during this period. Neurological evaluations performed subsequently showed that significantly greater deficits were seen in the groups receiving thiopental or placebo than in the group receiving phenytoin at 15 and 30 min after tourniquet release. This same trend, though of lesser magnitude, was again observed at 72 hours. Histological examination of the brains, obtained 72 hours postischemia, showed necrotic injuries occurring in 42% of the animals receiving placebo, 8% of those given thiopental, and in none of the phenytoin group (p < 0.01). Evidence of ischemic lesions was noted in 33, 42, and 33% of these same groups, respectively. Pretreatment with phenytoin appeared to protect against ischemic cerebral lesions, as evidenced by both neurological recovery and postmortem findings.