This review updates the molecular pathological classification of colorectal cancer (CRC), highlighting recent advancements in bulk and single-cell RNA data analysis for transcriptional classifiers and risk stratification. CRC is classified into four consensus molecular subtypes (CMS) based on bulk RNA data, with significant inter-tumoural and intra-tumoural heterogeneity. The tumour microenvironment (TME) plays a crucial role in prognosis, and efforts to isolate neoplastic-specific traits have led to the identification of CRC intrinsic subtypes (CRIS). Immunohistochemistry and digital pathology are evolving fields in CRC classification. Key properties include conventional adenoma versus serrated polyp pathway transcriptomic analysis and characterisation of canonical LGR5+ crypt base columnar stem cells versus ANXA1+ regenerative stem cells. Three pathway-derived subtypes (PDS1-3) have been developed, revealing a continuum of intrinsic biology associated with biological, stem cell, histopathological, and clinical attributes. The CMS classification, developed in 2015, uses bulk RNA sequences to classify CRC into four major groups. The CRIS classification, published in 2016, separated CRC epithelial neoplastic cells into five subtypes. The single-cell intrinsic consensus molecular subtype (iCMS) classification, published in 2022, identified two intrinsic subtypes, iCMS2 and iCMS3. The PDS classification, published in 2023, splits CRC into three subtypes based on Gene Ontology inferred pathway activation patterns. The role of the stroma in transcriptome and prognosis has been highlighted, with studies showing that fibroblast infiltration and genes from cancer-associated fibroblasts are key factors in disease relapse. The CRIS approach identified five new subtypes, CRIS-A to CRIS-E. The CMS classification has been used in clinical samples to enable retrospective alignment with outcome and treatment response. The role of the stroma in CRC classification has been clearly defined by Jass et al. in the late 1980s. The promise of precision medicine in CRC has been heralded by the development of CMS and other tools, but the absence of large clinical impact may be seen as a failure over the last decade. Rapid turnaround CMS classification and the emergence of morphology, immunohistochemistry (IHC), and image-based surrogates have been discussed. A five-marker IHC panel was developed to deliver a practical classification tool with 87% concordance with the 'gold-standard' transcriptomic CMS classification. Digital pathology and image-based H&E approaches have enabled the development of image-based CMS (imCMS) deep learning classifiers. The single-cell intrinsic CMS (iCMS) classifier identified two epithelial classes with distinct gene expression, transcriptional factor activity, and genomic profiles. The iCMS2 class was associated with SCNA/copy number variation (CNV), whereas iCMS3 displayed limited uniformityThis review updates the molecular pathological classification of colorectal cancer (CRC), highlighting recent advancements in bulk and single-cell RNA data analysis for transcriptional classifiers and risk stratification. CRC is classified into four consensus molecular subtypes (CMS) based on bulk RNA data, with significant inter-tumoural and intra-tumoural heterogeneity. The tumour microenvironment (TME) plays a crucial role in prognosis, and efforts to isolate neoplastic-specific traits have led to the identification of CRC intrinsic subtypes (CRIS). Immunohistochemistry and digital pathology are evolving fields in CRC classification. Key properties include conventional adenoma versus serrated polyp pathway transcriptomic analysis and characterisation of canonical LGR5+ crypt base columnar stem cells versus ANXA1+ regenerative stem cells. Three pathway-derived subtypes (PDS1-3) have been developed, revealing a continuum of intrinsic biology associated with biological, stem cell, histopathological, and clinical attributes. The CMS classification, developed in 2015, uses bulk RNA sequences to classify CRC into four major groups. The CRIS classification, published in 2016, separated CRC epithelial neoplastic cells into five subtypes. The single-cell intrinsic consensus molecular subtype (iCMS) classification, published in 2022, identified two intrinsic subtypes, iCMS2 and iCMS3. The PDS classification, published in 2023, splits CRC into three subtypes based on Gene Ontology inferred pathway activation patterns. The role of the stroma in transcriptome and prognosis has been highlighted, with studies showing that fibroblast infiltration and genes from cancer-associated fibroblasts are key factors in disease relapse. The CRIS approach identified five new subtypes, CRIS-A to CRIS-E. The CMS classification has been used in clinical samples to enable retrospective alignment with outcome and treatment response. The role of the stroma in CRC classification has been clearly defined by Jass et al. in the late 1980s. The promise of precision medicine in CRC has been heralded by the development of CMS and other tools, but the absence of large clinical impact may be seen as a failure over the last decade. Rapid turnaround CMS classification and the emergence of morphology, immunohistochemistry (IHC), and image-based surrogates have been discussed. A five-marker IHC panel was developed to deliver a practical classification tool with 87% concordance with the 'gold-standard' transcriptomic CMS classification. Digital pathology and image-based H&E approaches have enabled the development of image-based CMS (imCMS) deep learning classifiers. The single-cell intrinsic CMS (iCMS) classifier identified two epithelial classes with distinct gene expression, transcriptional factor activity, and genomic profiles. The iCMS2 class was associated with SCNA/copy number variation (CNV), whereas iCMS3 displayed limited uniformity