Background: Targeted therapeutic strategies for advanced colorectal cancer (CRC) have been limited.STING is crucial to the antitumor immunotherapy, for it stimulates IFN signaling to mediate the crosstalk between innate and adaptive immune responses. Emerging evidence suggests that STING also contributes to the prognosis of CRC. However, prognostic models relating to STING have not yet been explored.Methods: A total of 431 CRC samples from the TCGA database were analyzed to explore the prognostic value of STING-related genes. We trained prognostic models using the multivariate Cox regression. A STING-related prognostic score (SPS) was calculated as the gene expression multiplied by the corresponding coefficients of the final model. A backward stepAIC strategy was adopted to select the optimal model. A nomogram was used to personalize medical decisions for CRC.Results: The expression level of STING was upregulated in the CMS1 subtype (P=0.036). Among STING-related genes, DHX9 (HR =0.72, P=0.01), IRF2 (HR =1.34, P=0.022), and POLR1D (HR =1.23, P=0.038) showed significant prognostic value. The SPS was proven to be an independent risk factor (training: HR =2.9, P=0.00013; validation: HR =3.02, P=0.01), and outperformed random classifiers in identifying high-risk CRC. The high SPS group was characterized by less genomic aberrations, upregulated IL6-JAK-STAT3 and IL2-STAT5 signaling pathways, increased expression of TIM-3, increased infiltration of regulatory T (Treg) cells and T helper 17 (Th17) cells, and decreased infiltration of M0 macrophages.Finally, the nomogram based on the SPS and clinical factors showed good performance in CRC.Conclusions: SPS is an independent risk factor that could identify high-risk CRC. While ICBs may benefit patients of the CMS1 subtype, for the CMS2, CMS3, and CMS4 subtypes in the high SPS group, STING agonists and immunotherapies targeting the Th17 axis may be beneficial. Finally, the SPS-based nomogram could help advance personalized medical decisions for CRC.