Colorectal cancer (CRC) exhibits significant molecular and immunological heterogeneity. Neutrophil infiltration patterns play a crucial yet poorly understood role in tumor progression and patient outcomes. This study presents a comprehensive single-cell atlas of the CRC tumor microenvironment (TME), integrating transcriptomic data from 388,511 cells across 98 samples from 63 patients. Employing advanced computational methods, we stratified patients based on their immune cell infiltration profiles, revealing distinct immunophenotypes with potential therapeutic implications. Our analysis focused on tissue-resident neutrophils (TRNs) and uncovered previously uncharacterized subpopulations with diverse functional states. Trajectory inference analysis revealed a dynamic differentiation path from normal-associated neutrophils to tumor-associated neutrophils, highlighting the remarkable plasticity of these cells within the tumor environment. By integrating single-cell data with bulk transcriptomic and clinical information, we identified specific neutrophil-derived gene signatures associated with poor prognosis in CRC, suggesting their potential as novel prognostic biomarkers. This study not only provides unprecedented insights into neutrophil heterogeneity in CRC but also identifies potential targets for immunomodulatory therapies. Our findings lay the groundwork for developing more nuanced, personalized immunotherapeutic strategies for CRC, potentially improving treatment efficacy for patients who currently show a limited response to existing immunotherapies.