Background: The prevalence of autoimmune hepatitis (AIH) is increasing, and its early clinical diagnosis is difficult. The pathogenesis of AIH remains unclear, and AIH-related studies are largely limited because of lack of suitable mouse models. Methods: To obtain a good tool for research on AIH, we first established an improved immune-mediated mouse model that can mimic the pathological process of AIH as in the human body, through repeated injections of human cytochrome P450 2D6 (CYP2D6) plasmid. Next, a proteomic analysis based on isobaric tag (IBT) technology was performed to detect the differentially expressed proteins (DEPs), and related biological functions and pathways in the plasma of AIH and normal mice. Finally, we performed enzyme-linked immunosorbent assay (ELISA) to further confirm the most abundant DEP in the plasma of patients with AIH. Results: Autoantibodies and the characteristic pathology of AIH were observed in our mouse model. Inflammatory infiltration also increased in the livers of AIH mice over time and plateaued by day 42 post the first injection. Chronic hepatitis was most severe on day 35 with the development of fibrosis as well, and the plasma of AIH mice were collected for proteomic analysis. A total of 176 DEPs were found in this experiment, of which 148 DEPs were up-regulated and 28 DEPs were down-regulated. Thirty significant Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways (P < 0.05) were detected. Arginine biosynthesis was found to be the most significant pathway involved in the AIH process. During the Gene Ontology (GO) analysis, most DEPs were found to be involved in the binding, cellular, and metabolic processes. Using ELISA, the most overexpressed DEP, serum amyloid A 1 (SAA1), was confirmed to be increased specifically in the plasma of patients with AIH compared to other chronic hepatitis. Different plasma levels of SAA1 were also found related to different grades of inflammation and stages of fibrosis in the liver of patients with AIH. Conclusions: Our study is the first to describe the proteomics analysis of a true sense of AIH mouse model, which is beneficial for a better understanding of AIH pathogenesis and identifying potential biomarkers for its clinical diagnosis.