There has been little systematic examination concerning the comparative effects of the anesthetized versus the awake state on outcome from cerebral ischemia. This experiment evaluated infarct volume and neurologic function in rats subjected to temporary focal ischemia while anesthetized with either sevoflurane or halothane. Outcome in these animals was compared to that observed in rats maintained unanesthetized during a similar ischemic insult. All rats were anesthetized with halothane and surgically prepared for filament occlusion of the middle cerebral artery. After preparation, one group (Halothane) remained anesthetized with approximately 1.4 MAC halothane. In another group (Sevoflurane), halothane was discontinued and substituted with sevoflurane, which was administered until electroencephalographic burst suppression was evident (approximately 1.4 MAC). The final group (Awake) was allowed to awaken immediately after the onset of ischemia. Middle cerebral artery occlusion persisted for 90 min in all groups. The middle cerebral artery filament then was removed, and a 96-h survival interval was allowed. Neurologic function and infarct volume were determined. Recent evidence indicates that transient mild hyperthermia occurs in awake rats undergoing filament occlusion of the middle cerebral artery. To examine the potential role of mild hyperthermia in this experiment, a second experiment was performed in which rats anesthetized with halothane underwent 90-min focal ischemia, with pericranial temperatures held at either 38.0 degrees C or 39.2 degrees C. Intraischemic mean arterial pressure was 20-25 mmHg lower in the two anesthetized groups compared with awake animals. Despite this finding, cortical infarct volumes (mean +/- SD; Halothane, 17 +/- 32 mm3; Sevoflurane, 36 +/- 57 mm3; Awake, 115 +/- 104 mm3; Sevoflurane, 36 +/- 57 mm3; Awake, 115 +/- 104 mm3) and subcortical infarct volumes (mean +/- SD; Halothane, 39 +/- 57 mm3; Sevoflurane, 50 +/- 29 mm3; Awake, 88 +/- 46 mm3) were reduced in both groups of anesthetized rats. This reduction correlated with improved neurologic function. The rats in whom the pericranial temperature was maintained at 39.2 degrees C had a larger total infarct volume (218 +/- 81 mm3) and increased neurologic deficits when compared to those in whom the pericranial temperature was maintained at 38.0 degrees C (total infarct volume, 75 +/- 77 mm3). Both halothane and sevoflurane substantially reduced damage in this focal ischemia model when compared to outcome resulting from the same insult induced in awake rats. The reduction in intraischemic mean arterial pressure caused by the anesthetics did not seem contributory to outcome. Brain temperature differences among the groups were not defined. Because small differences in pericranial temperature were shown to have major effects on outcome, further work is required to determine if differences in brain temperature explain the observed protective effects of these anesthetics.