Using a rat model of incomplete cerebral ischemia the effects of isoflurane (iso) and methohexital (metho) were compared with those of 70% nitrous oxide controls (N2O). Two levels of incomplete cerebral ischemia were produced by right carotid occlusion plus hypotension for 30 min: moderate = 30 mmHg, FIo2= 0.30; severe = 25 mmHg, FIo2= 0.20. The iso doses (1 and 2 MAC) and metho doses (0.01 and 0.1 mg·kg−1·min−1) were tested at each ischemic level. These iso and metho doses were selected because without ischemia they produced similar decreases in cerebral oxygen consumption (CMRo2) compared with that produced in N2O controls. In the absence of ischemia, the electroencephalogram (EEG) was suppressed by 0.01 mg·kg−1·min−1metho and 1 MAC iso and showed burstsuppression with 0.1 mg·kg−1·min−1metho and 2 MAC iso. The EEG was further depressed by ischemia under all anesthetic conditions. Neurologic outcome was evaluated for 3 days following incomplete cerebral ischemia by using a graded deficit score (0 = normal, 5 = death associated with stroke). Following moderate ischemia all four anesthetic treatments improved outcome compared with N2O controls, but after severe ischemia only 2 MAC iso significantly improved outcome. Neurohistopathology was evaluated on a scale of 0 to 40, 24 h after ischemia. The neurohistopathology score was significantly improved by all four anesthetic treatments compared with N2O following moderate ischemia and was better with 2 MAC iso compared with 0.1 mg·kg−1·min−1metho after both moderate and severe ischemia. These results show that both iso and metho improve outcome from cerebral ischemia compared with that associated with N2O, but only 2 MAC iso resulted in an improved outcome following severe ischemia. This difference in outcome between the two anesthetics may be related to greater neuronal depression with iso, which may occur with little difference in cerebral metabolic depression.