Colorectal cancer is a very common cancer worldwide. Serum tumor-associated autoantibodies (TAAbs), especially the anti-p53 autoantibody, may be promising biomarkers to detect early-stage colorectal cancer. This study aimed to identify all known autoantibodies and their value in colorectal cancer diagnosis, as well as exploring the underlying connections and mechanisms through a bioinformatics analysis. Databases were used to select available articles of TAAbs in colorectal cancer. In a meta-analysis of the anti-p53 autoantibody, the diagnostic odds ratio and area under the curve (AUC) of the summary receiver-operating characteristic (SROC) curve were calculated using Stata 12.0 and Meta-Disc 1.4. We identified 73 articles including 199 single autoantibodies and 42 multiple autoantibodies. The maximum value of Youden's index was 0.76, combining c-MYC, p53, cyclin B1, p62, Koc, IMP1, and survivin. The diagnostic odds ratio for anti-p53 autoantibody at all stages was 10.86 (95% CI 8.40, 14.06) with low heterogeneity (I 2 = 40.3%) and the AUC of the SROC curve was 0.82. For the anti-p53 autoantibody in early-stage colorectal cancer, the diagnostic odds ratio was 4.82 (95% CI 2.95, 7.87) with heterogeneity (I 2 = 7.9%) and the AUC of the SROC curve was 0.72. Eighty-seven autoantibodies were selected for bioinformatics analyses. We found that the most enriched functional terms and protein-protein interactions may relate to the mechanism of autoantibody generation. In summary, our study summarized the diagnostic value of TAAbs in colorectal cancer, either as single molecules or in combination. Bioinformatics analyses may be a new approach to explore the mechanism of autoantibody generation.