This study introduces a novel approach to molecular subtyping in colorectal cancer (CRC) using pathway-level data, which identifies three distinct pathway-derived subtypes (PDS1, PDS2, and PDS3). PDS1 tumors are characterized by canonical/LGR5+ stem-rich, highly proliferative traits and good prognosis. PDS2 tumors are regenerative/ANXA1+ stem-rich, with elevated stromal and immune tumor microenvironmental lineages. PDS3 tumors, a previously overlooked subset within CMS2, have reduced stem populations and increased differentiated lineages, particularly enterocytes and enteroendocrine cells, but display the worst prognosis in locally advanced disease. These PDS3 phenotypic traits are evident across multiple datasets and highlight a series of subtle biological states that are underrepresented in pre-clinical models. The study also demonstrates the clinical relevance of PDS classification, with PDS3 tumors showing poor relapse-free survival rates compared to PDS1 and PDS2 tumors. Additionally, the lack of pre-clinical models for PDS3 tumors limits the ability to study their biology and therapeutic options. The findings suggest that combining PDS with existing subtyping approaches can provide a more comprehensive understanding of CRC biology and improve clinical outcomes.This study introduces a novel approach to molecular subtyping in colorectal cancer (CRC) using pathway-level data, which identifies three distinct pathway-derived subtypes (PDS1, PDS2, and PDS3). PDS1 tumors are characterized by canonical/LGR5+ stem-rich, highly proliferative traits and good prognosis. PDS2 tumors are regenerative/ANXA1+ stem-rich, with elevated stromal and immune tumor microenvironmental lineages. PDS3 tumors, a previously overlooked subset within CMS2, have reduced stem populations and increased differentiated lineages, particularly enterocytes and enteroendocrine cells, but display the worst prognosis in locally advanced disease. These PDS3 phenotypic traits are evident across multiple datasets and highlight a series of subtle biological states that are underrepresented in pre-clinical models. The study also demonstrates the clinical relevance of PDS classification, with PDS3 tumors showing poor relapse-free survival rates compared to PDS1 and PDS2 tumors. Additionally, the lack of pre-clinical models for PDS3 tumors limits the ability to study their biology and therapeutic options. The findings suggest that combining PDS with existing subtyping approaches can provide a more comprehensive understanding of CRC biology and improve clinical outcomes.