Complete global ischemia was produced in 39 dogs by temporary ligation of the aorta. Prior to the ischemic episode, pentobarbital (30 to 45 mg per kilogram of body weight) was administered to 19 of these dogs. The neurological effects of cerebral ischemia episodes lasting 8, 9, or 10 minutes were compared in dogs treated with pentobarbital and those not treated. At 48 hours following the ischemic episode most of the dogs made ischemic for 8 minutes were normal, whereas most animals made ischemic for 10 minutes were dead or comatose. The 9-minute ischemic period resulted in a relatively even distribution of normal and damaged dogs. There were no differences between treated and untreated dogs. Cerebral blood flow, cerebral metabolic rate for oxygen, and various cerebral metabolites were measured in dogs surviving 48 hours. Again, there were no differences between treated and untreated dogs. We conclude that barbiturates provide no protection in this model of complete global ischemia. This conclusion supports the hypothesis that the likely mechanism of barbiturate protection in models of incomplete ischemia or hypoxia is based on cerebral metabolic depression; such a mechanism would not be expected to be effective in complete global ischemia.