Abstract Pancreatic ductal adenocarcinoma (PDAC) is notable due to its complex and dense stroma which comprises a high proportion of non-malignant cells. Studying PDAC epithelial cells at the single cell level may yield insights which have not been appreciated through bulk tissue analysis. We analysed a novel single cell RNA sequencing (scRNA seq) dataset of 11000 epithelial (EPCAM+) cells from 7 resected PDAC samples and used a validation cohort of 43 samples from two published scRNA seq datasets. Malignant epithelial cells were identified by: gene expression, changes in copy number, and by the presence of known PDAC single nucleotide variants. This confirmed that sorting for EPCAM+ cells prior to scRNA seq resulted in an enriched proportion of malignant vs non-malignant tumor epithelial cells. This dataset of enriched single cell dataset was therefore useful for studying the influence of intrinsic malignant cell factors on PDAC biology. We observed that malignant cells do not segregate into classical or basal-like signatures, despite these being clinically validated bulk transcriptomic tumor subtypes. Indeed, classical and basal-like signatures module scores in PDAC in individual malignant cells were positively correlated, suggesting they are not divergent programs in malignant cells. Bulk transcriptomic signature subtypes are therefore most likely to reflect differences in PDAC tissue stroma, rather than in intrinsic malignant cell factors. Moreover, we sort to investigate intra- and inter- sample heterogeneity to understand the intrinsic malignant cell factors which may influence the biology of PDAC. We analysed single cell velocity to understand the dynamics of malignant cell states. This showed heterogeneity in expression of known pan-cancer malignant cell meta-programs which identified a population of cells with high expression of cell cycle genes, suggesting greater proliferation. We then explored expression of immune modulating genes to determine how malignant cells influence the immunosuppressive microenvironment. For example, we identified heterogeneity in HLA class 1 and antigen presentation gene expression which may be a mechanism of immune suppression in PDAC. Together these data demonstrate how studying scRNA seq of EPCAM+ cells reveals intrinsic malignant cell factors which cannot be appreciated through bulk RNA seq. Citation Format: Edward H. Arbe-Barnes, Ashwin Jainarayanan, Enas Abu-Shah, Rachael J.M. Bashford-Rogers, Shivan Sivakumar. Single cell RNA sequencing of EPCAM+ cells reveals unique malignant biology [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Pancreatic Cancer; 2023 Sep 27-30; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(2 Suppl):Abstract nr A106.