Objective To compare the condition of illness and pathological characteristics of experimental autoimmune encephalomyelitis (EAE)in C57 BL/6 mouse models induced by myelin oligodendrocyte glycoprotein 35-55 (MOG35-55)at different doses,and provide a reliable animal model for further study of multiple sclerosis(MS).Methods Male SPF-grade C57 BL/6 mice were divided randomly into four groups:normal group and three EAE model groups (MOG35-55 high-dose,middle-dose and low-dose model groups).200,100,50μg MOG35-55/mice were mixed with complete Freund's adjuvant(CFA),respectively,to prepare complete antigen in different concentrations.The mice were anesthetized and injected s.c.over flanks with the complete antigen and injected i.P.with pertussis toxin to establish immunization-induced C57BL/6 mouse-model of EAE.The mice of the normal group were injected with normal saline instead.Since the day of immunization,the incidence,body weight and neurological score of the mice were observed.The mice of different neurological scores in different periods were anesthetized and perfused with saline and followed by 4% paraformaldehyde.The brain and spinal cord of the mice were removed and fixed in the same fixative solution.The brains and spinal cords of the mice were examined by histopathology with hematoxylin-eosin(HE) staining.The mice on the 40th day were sacrificed and perfused with 2% paraformaldehyde and 2% glutaraldehyde, 1 mm~3 pieces of cerebral white matter and intumescentia lumbalis of the spinal cord were taken and ultrathin sections were prepared according to conventional techniques for electron microscopy. Results All the MOG_(35-55) in three different doses induced mouse models of EAE. The disease was with an incidence rate of 100% and a chronic monophasic course. The body weight of the mice in the three groups decreased obviously compared with those in the normal group. The maximum value of neurological score was 1.33,2.25 and 2.50 in the mice of high-, middle-and low-dose groups, respectively. The major histopathological changes observed in the brain and spinal cord of the EAE mice were different degrees of inflammatory cell infiltration around small vessels showing sleeve-like changes, dcmyelination and neuronal karyopyknosis in the acute and remission stages. The main site of the brain inflammation was in white matter around encephalocoele, and also in the DG and CA zones of hippocampus. The spinal cord inflammation was most severe in the lumbosacral region. The above mentioned pathological changes in the low-dose group were more prominent than those in the middle-dose and high-dose groups. The major ultrastructural changes were scattered around encephalocoele, interstitial edema, especially around small blood vessels, and swollen mitochondria with damaged cristae, and some karyopyknosis in vascular endothelial cells. Some tight junctions were blurred. Some dispersed lymphocytes and mononuclear cells were seen in the perivascular space. In lumbar intumescentia of the spinal cord, there were some myelin figures in the white matter myelin sheath. Some of them showed demyelization and structurtal fusion. The cytoplasmic organelles of axons were considerably reduced or even disappeared. The vascular basement membrane showed an increased thickness and focal necrosis in some areas. Conclusion The mouse models of immune-induced EAE are successfully established with MOG_(35-55), especially that induced with MOG in a dose of 50 μg. This mouse model is stable, with a high incidence and low mortality rate, and can be applied for EAE research in the future.