The purpose of this investigation was to determine the effects of sevoflurane on rats with ischemic brain injury and to determine the potential role of the TREK-1 channel in this process. Normal rats were randomly divided into three groups: Sham operation, sevoflurane anesthesia or chloral hydrate anesthesia group, an additional group of TREK-1 knockout rats were also studied. Semi-quantitative PCR and western blot analysis confirmed the lack of TREK-1 expression in the brain of TREK-1 knockout rats. The thread-tie method was used to establish middle cerebral artery occlusion (MCAO) model to induce cerebral ischemic brain injury. All rates were treated for 4 days prior to ischemia (for 2 h) followed by a 24 h reperfusion period. Physiological indexes of rats in each group both prior to and after surgery showed no statistical difference (P>0.05). Neurological function was scored both before (no statistical difference) and after surgery where it was found to be significantly better (lower score) in the sevoflurane anesthesia group than in chloral hydrate anesthesia and TREK-1 knockout groups (P<0.01). The area of cerebral infarction was measured by triphenyl tetrazolium chloride staining and terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay to detect the apoptosis of brain cells. TTC staining showed different degrees of cerebral infarction in the various groups; the area of cerebral infarction in sevoflurane anesthesia group was significantly lower than that in chloral hydrate anesthesia and TREK-1 knockout groups (P<0.01). TUNEL assay showed that the number of TUNEL-positive cells was significantly lower in sevoflurane anesthesia group than in TREK-1 knockout and chloral hydrate anesthesia groups (P<0.01). In conclusion, results from this investigation showed that sevoflurane can protect the nerve function of rats with cerebral ischemic brain injury possibly by affecting the expression of proteins involved in the TREK-1 signaling pathway. © 2017, Spandidos Publications. All rights reserved.