Previous studies have shown that dexmedetomidine exerted anti-inflammatory effect on several animal models with inflammation, but the mechanism is not clear. This study intends to elucidate the anti-inflammatory mechanism of dexmedetomidine through the cholinergic anti-inflammatory pathway. To investigate this therapeutic potential of dexmedetomidine, a murine model of endotoxemia was established induced by lipopolysaccharide (LPS). Animals were assigned to one of four protocols. Protocol one: animals were randomly assigned to control group, dexmedetomidine group, and sterile saline group (n=20 each), and these animals were used for survival analysis. The survival rate was assessed up to 120 h after endotoxin injection. Protocol two: animals were randomly assigned to one of four groups (n=16 each): group 1 (group Saline), treated with sterile saline 15 min prior to endotoxin treatment (10 mg kg(-1) over 2 min); group 2 (group Dex), treated with dexmedetomidine 15 min prior to endotoxin treatment; group 3 (group αBGT+Dex), treated with alpha-7 nicotinic acetylcholine receptors (α7nAChR) antagonist alpha-bungarotoxin (αBGT, 1 μg/kg) 15 min prior to dexmedetomidine treatment; group 4 (group saline+Dex), treated with equivalent sterile saline 15 min prior to dexmedetomidine treatment. Protocol three: animals were randomly assigned to one of two groups (n=16 each): vagotomy group (group VNX+Dex), right cervical vagus nerve was exposed and transected; sham-operated group (group SHAM+Dex), the cervical vagus nerve was visualized, but was neither isolated from the surrounding tissues nor transected. Protocol four: animals were treated with dexmedetomidine (40 μg/kg) and sterile saline to observe the discharge activity of cervical vagus nerves by using BL-420F data acquisition and analysis system (n=16 each). In the survival analysis groups, the survival rate of dexmedetomidine group was significantly higher than that of the endotoxemia group (65 versus 25 %, P<0.01). Preemptive administration of dexmedetomidine significantly attenuated the cytokine response after lipopolysaccharide (LPS) induced endotoxemia (TNF-alpha, IL-1beta, IL-6, P<0.01, respectively). However, preemptive administration of dexmedetomidine failed to suppress cytokine response in α-bungarotoxin group and vagotomy group (TNF-alpha, IL-1beta, IL-6, P>0.05, respectively). Furthermore, preemptive administration of dexmedetomidine significantly increased the discharge frequency of cervical vagus nerves in comparison with sterile saline treatment (P<0.01).Our results demonstrate that the preemptive administration of dexmedetomidine increases the activity of cervical vagus nerve and have the ability to successfully improve survival in experimental endotoxemia by inhibiting the inflammatory cytokines release. However, administration of dexmedetomidine to vagotomy or α7 nAChR antagonist pretreatment mice failed to suppress TNF levels, indicating that the vagus nerve and α7nAChR-mediated cholinergic anti-inflammatory pathway is required for the anti-inflammatory effect of dexmedetomidine. These findings show that central alpha-2 agonist dexmedetomidine suppresses systemic inflammation through vagal- and α7nAChR-dependent mechanism.