Ｏｂｊｅｃｔｉｖｅ　To investigate the effects and possible mechanism of different doses dexmedetomidine(Dex) pretreatment on lidocaine toxicity to nervus centralis of albino rabbit.　Methods　Forty New-Zealand albino rabbit were randomly divided into 4 groups (n=10). Group A and B were received infusion of 8 ml mixture Dex (respectively, 5 μg/kg and 10 μg/kg) and saline，10 min later lidocaine was pumped at the rate of 4 mg·kg-1·min-1 until occurrence of hyperspasmia. The equal dosis of saline and lidocaine were given in group C, but group D were only received saline. The hematoplasma density of lidocaine was measure when hyperspasmia occurred, the doses of lidocaine, the appearing times of lidocaine-induced hyperspasmia were observed, the concentration of aspartate, glutamic acid, glycine and γ-aminobutyric acid in brain architecture were recorded.　Results　Compared with group C[(137±37) mg，(5.4±0.6) μg/kg, (510±76) s, respectively], the doses of lidocaine when neurotoxicity occur, the plasmic density of lidocaine and the appearing times of lidocaine-induced hyperspasmia were all increased obviously in group A [(240±48) mg，(6.4±0.8) μg/kg，(822±122) s, respectively] and B[(230±51) mg，(6.3±0.5) μg/kg，(802±114) s, respectively] (P<0.05). Compared with group D [(1.5±0.8), (2.4±1.2), (4.7±1.6), (5.7±2.8) μmol/g, respectively], the concentration of aspartate, glutamic acid, glycine and gamma-aminobutyric acid in brain architecture were all increased in group A[(3.5±1.0), (4.0±1.9), (10.1±1.9), (16.5±2.2) μmol/g, respectively], B[(3.7±0.8), (4.2±1.9), (11.4±2.2), (17.4±2.4) μmol/g, respectively] and C[(4.7±1.0), (6.8±1.9), (13.7±1.9), (20.9±3.4) μmol/g, respectively](P<0.05), then in group C obviously higher than in group A and B (P<0.05). Conclusions　 Dex pretreatment may postpone the occurrence of lidocaine-induced hyperspasmia and raise the neurotoxic threshold of lidocaine, so it has protection on nervus centralis.