The need for reliable biomarkers for distinguishing Crohn disease (CD) from ulcerative colitis (UC) is increasing. This study aimed at evaluating the diagnostic potential of anti-GP2 antibodies as a biomarker in Chinese patients with CD. In addition, a variety of autoantibodies, including anti-saccharomyces cerevisiae antibodies (ASCA), perinuclear anti-neutrophil cytoplasmic antibodies (PANCA), anti-intestinal goblet cell autoantibodies (GAB), and anti-pancreatic autoantibodies (PAB), were evaluated.A total of 91 subjects were prospectively enrolled in this study, including 35 patients with CD, 35 patients with UC, 13 patients with non-IBD gastrointestinal diseases as disease controls (non-IBD DC), and 8 healthy controls (HC). The diagnosis of IBD was determined based on the Lennard-Jones criteria, and the clinical phenotypes of the IBD patients were determined based on the Montreal Classification.Anti-GP2 IgG antibodies were significantly elevated in patients with CD, compared with patients with UC (P = 0.0038), HC (P = 0.0055), and non-IBD DC (P = 0.0063). The prevalence of anti-GP2 IgG, anti-GP2 IgA and anti-GP2 IgA, or IgG antibodies in patients with CD was 40.0%, 37.1%, and 54.3%, respectively, which were higher than those in non-IBD DC (anti-GP2 IgG, 15.4%; anti-GP2 IgA, 7.7%; and anti-GP2 IgA or IgG, 23.1%) and those in patients with UC (anti-GP2 IgG, 11.4%; anti-GP2 IgA, 2.9%; and anti-GP2 IgA or IgG, 14.3%). For distinguishing CD from UC, the sensitivity, specificity, positive predictive value (PPV) and positive likelihood ratios (LR+) were 40%, 88.6%, 77.8%, and 3.51 for anti-GP2 IgG, 37.1%, 97.1%, 92.9%, and 13.0 for anti-GP2 IgA, and 54.3%, 85.3%, 79.2%, and 3.69 for anti-GP2 IgA or IgG. For CD diagnosis, the combination of anti-GP2 antibodies with ASCA IgA increased the sensitivity to 68.6% with moderate loss of specificity to 74.3%. Spearman's rank of order revealed a significantly positive correlation of anti-GP2 IgG with ileocolonic location of disease (L3) (P = 0.043) and a negative correlation of anti-GP2 IgA with biologic therapy (P = 0.012).Our findings suggest that anti-GP2 antibodies could serve as a biomarker for distinguishing patients with CD from patients with UC, and the combination of anti-GP2 antibodies with ASCA IgA may improve the predictive power.