Abstract: Objective To establish the rabbits model of selectively anesthetized brain and spinal cord and to explore the skeletal muscle relaxation sites of sevoflurane. Methods Sixteen adult male New Zealand white rabbits were randomly assigned to the experiment. The upper torso systemic circulation and the lower torso bypass circulation were independently established by the ligation of thoracic aorta at T12-L1 level. Sevoflurane was administered to the upper or lower torso through lungs or oxygenator to selectively anesthetized brain or spinal cord (mainly lumbar and sacro-coccygeal region). Sevoflurane concentration from end-tidal (represented the brain) and oxygenator outlet (represented the spinal cord) was measured by an anesthetic gas analyzer. The concentration and partial pressure of sevoflurane in carotid artery (represented the brain) and abdominal aorta (represented the spinal cord) were determined using a gas chromatograph with the two-stage headspace equilibration method. Results When 1.5 mininum alveolar concentration （MAC） sevoflurane was administrated via lungs, the concentration and partial pressure of sevoflurane in the carotid artery were significantly higher than those in the abdominal aorta （P<0.05）, with the end-tidal sevoflurane concentration higher than that of oxygenator outlet （P<0.05), which indicated sevoflurane concentration and partial pressure in brain were higher than those in spinal cord. When 1.5 MAC sevoflurane was delivered via oxygenator,the indicators were conversed （P<0.05）, which indicated sevoflurane concentration and partial pressure in spinal cord were higher than those in brain. Conclusion Based on the unique blood supply to the spinal cord of New Zealand white rabbits, we successfully established selectively anesthetized brain and spinal cord rabbit models.