Colorectal cancer (CRC) is the third most common malignant tumor. DNA damage played a crucial role on tumorigenesis, and abnormal DNA repair pathways affected the occurrence and progress of CRC. In current study, we aimed to construct a DNA repair-related genes (DRG) signature to predict the overall survival (OS) of CRC patients. Differentially expressed DRGs (DE-DRGs) were analyzed using The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. The prognostic gene signature was identified by univariate Cox regression and Least Absolute Shrinkage and Selection Operator (LASSO)-penalized Cox proportional hazards regression analysis. The predictive ability of the model was evaluated by utilizing Kaplan-Meier curves and Time-dependent receiver operating characteristic (ROC) curves. Gene set enrichment analysis (GSEA) was performed to explore the underlying biological processes and signaling pathways. Single sample gene set enrichment analysis (ssGSEA) was implemented to estimate the immune status between the different risk group. A total of 118 DE-DRGs were identified between 42 normal samples and 488 CRC samples in TCGA cohort, and 36 DE-DRGs were associated with OS in the univariate Cox regression. A 9 DE-DRGs (ESCO2, AXIN2, PLK1, CDC25C, IGF1, TREX2, ALKBH2, ESR1 and MC1R) signature was constructed to classify patients into high-risk and low-risk group. The risk score was an independent prognostic indicator for OS (HR>1, P<0.001). Genetic alterations analysis indicated that the 9 DE-DRGs in the signature were changed in 63 required samples (100%) and the major alteration was missense mutation. The function enrichment analysis indicated that RNAi effector complex, sensory perception of smell and olfactory receptor activity were the main biological processes. The high-risk group had higher immune cell infiltration levels than low-risk group. The 9 DE-DRGs signature was significantly associated with OS and provided a new insight for the diagnosis and treatment for CRC. The predictive ability of this model need be supported by further basic researches.