Background: As the most common gastrointestinal malignancy worldwide, liver metastases occur in half colorectal cancer (CRC) patients. Early detection can help treat them early and reduce mortality in patients with colorectal cancer liver metastases (CRLM). Finding useful biomarkers for CRLM is thus essential., Methods: The TCGA and GEO databases were used to download the expression profiles and clinical data of the patients. Differential analysis screened for genes associated with CRLM, and univariate Cox regression analysis identified genes associated with prognosis. The least absolute shrinkage and selection operator (LASSO) method further preferred genes to construct a prognostic signature. Kaplan-Meier survival curves were used to show patients' overall survival (OS). Receiver operating characteristic (ROC) curves showed the accuracy of the model. Risk scores and clinical characteristics of patients were included in multivariate Cox regression analysis to identify independent risk factors, and a nomogram was constructed. The proportion of immune cells and infiltration were assessed using the 'CIBERSORT' package and the 'ESTIMATE' package., Results: We constructed a signature consisting of seven CRLM-associated genes, and signature-based risk scores have great potential in estimating the prognosis of CRC patients. Moreover, the poor response to immunotherapy in high-risk patients might contribute to the poor prognosis of individuals. Furthermore, we found that overexpression of Hepcidin antimicrobial peptide (HAMP), the only gene highly expressed in CRC and liver metastatic tissues, promoted CRC development and that it was associated with tumor mutation burden (TMB), DNA mismatch repair (MMR) genes, and microsatellite instability (MSI) in various tumors. Finally, we found that in CRC patients, low expression of HAMP also represented a better immunotherapeutic outcome, reflecting the critical role of HAMP in guiding immunotherapy., Conclusion: We identified a prognostic signature containing 7 CRLM-associated genes, and the signature was specified as an independent predictor and a nomogram containing the risk score was built accordingly. In addition, the derived gene HAMP could help guide the exploration of profitable immunotherapeutic strategies., Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2023 Liu, Lu, Yan, Xue, He, Huang, Sun, Zhao and Li.)