ObjectiveTo investigate the possible mechanism of electroacupuncture at Fengchi （GB 20）， Yanglingquan （GB 34） and Waiguan （TE 5） for acute migraine attacks by peroxisome proliferator-activated receptor γ/ nuclear transcription factor-κB pathway in the trigeminal neurovascular system. MethodsForty SD rats were randomly divided into blank group， model group， electroacupuncture group， electroacupuncture plus inhibitor group， and agonist group， 8 rats in each group. Except for the blank group， rats were injected with inflammation decoction intracranial catheter to establish rat models of acute migraine. Thirty minutes after modelling， the electroacupuncture group was given 0.9% NaCl solution by intraperitoneal injection and then electroacupuncture at "Fengchi" （GB 20）， "Yanglingquan" （GB 34）， and "Waiguan" （TE 5） for 30 mins； the electroacupuncture plus inhibitor group was given PPARγ inhibitor T0070907 （1.5 mg/kg） and eclectroacupuncture as the above for 30 mins； the agonist group was given PPARγ inhibitor pioglitazone hydrochloride （10 mg/kg） for 30 mins. Rats in the blank group and the model group were injected intraperitoneally with 0.9% NaCl solution and then bound and restrained for 30 mins. Behavioural scores were evaluated before modelling， 30 mins after modelling （pre-intervention） and post-intervention； after the last behavioural test， PPARγ， NF-κB p65 mRNA content in rat dura mater was detected by real-time fluorescence quantitative PCR； protein expression of PPARγ， NF-κB p65， interleukin 6 （IL-6）， tumour necrosis factor α （TNF-α） in spinal trigeminal nucleus of rats was detected by protein blotting； immunofluorescence double-labelling method was used to detect the fluorescence intensity of PPARγ， NF-κB p65 in spinal trigeminal nucleus of rats. ResultsCompared with the blank group at the same time， the behavioural scores of rats in the remaining groups increased after modelling and after intervention （P＜0.01）. Compared with the model group at the same time， the beha-vioural scores of rats in the electroacupuncture group， electroacupuncture plus inhibitor group， and agonist group decreased after the intervention （P＜0.01）. Compared with the electroacupuncture group at the same time， beha-vioural scores reduced in the agonist group and elevated in the electroacupuncture plus inhibitor group after intervention （P＜0.01）. Compared with the blank group， expression of PPARγ and NF-κB p65 mRNA elevated in the dura mater of rats in the model group， and expression of PPARγ， NF-κB p65， IL-6， and TNF-α protein enhanced in spinal trigeminal nucleus， and the fluorescence intensity of PPARγ and NF-κB p65 enhanced （P＜0.01）. Compared with the model group， PPARγ mRNA expression in dura mater elevated and NF-κB p65 mRNA expression reduced in each intervention group， PPARγ protein expression in spinal trigeminal nucleus enhanced， and NF-κB p65， IL-6， and TNF-α protein expression weakened； in the electroacupuncture group and the agonist group， PPARγ fluorescence intensity enhanced， and the fluorescence intensity of NF-κB p65 weakened in each intervention group （P＜0.05 or P＜0.01）. Compared with the electroacupuncture group， in the electroacupuncture plus inhibitor group， PPARγ mRNA， protein expression and fluorescence intensity reduced， NF-κB p65 mRNA， protein expression and fluorescence intensity elevated， and IL-6 and TNF-α protein expression enhanced； in the agonist group， PPARγ mRNA and protein expression elevated， NF-κB p65 mRNA and protein expression reduced， and IL-6 and TNF-α protein expression decreased （P＜0.05 or P＜0.01）. ConclusionElectroacupuncture at "Fengchi" （GB 20）， "Yanglingquan" （GB 34） and "Waiguan" （TE 5） can can reduce the symptoms of acute migraine attacks in rats， and its mechanism may be related to the up-regulation of PPARγ expression and the reduction of NF-κB expression， thus inhibiting the neurogenic inflammatory response.