Objective <p>This study aims to construct a prognostic model based on neutrophil heterogeneity by integrating single-cell and bulk transcriptomic data. Additionally, it seeks to explore the features of the immune microenvironment across different risk subgroups to facilitate precise prognostic assessment and individualized immunotherapy for colorectal cancer.</p> Methods <p>Colorectal cancer transcriptomic datasets from TCGA and GEO were utilized. Single-cell data underwent quality control, clustering, and cell annotation. Cell communication analysis characterized neutrophil signaling interactions. Candidate genes were identified through differential expression analysis and WGCNA, leading to the establishment and validation of a CCL15/CCDC154 dual-gene prognostic model. Differences in immune infiltration, immune checkpoint expression, and immune escape between high- and low-risk groups were analyzed, followed by molecular subtype clustering.</p> Results <p>Significant remodeling of the tumor microenvironment was observed, with neutrophils extensively participating in multicellular signaling networks. The dual-gene model demonstrated strong predictive performance and functioned as an independent prognostic indicator. The high-risk group exhibited a pronounced immunosuppressive phenotype, characterized by significant immune escape and upregulated inhibitory immune checkpoints. Notably, distinct immune and prognostic differences were identified among molecular subtypes.</p> Conclusion <p>The dual-gene model based on neutrophil heterogeneity facilitates accurate prognostic stratification. The neutrophil-mediated immune microenvironment is closely linked to clinical outcomes, providing a theoretical basis for personalized treatment strategies in colorectal cancer.</p>