Pancreatic cancer is an important cause of morbidity and mortality but there are presently few if any prognostic markers available beyond tumour stage and grade. As tumours are detected and treated at earlier stages these factors are less prognostically informative and there is a growing need for development of novel diagnostic markers to guide pre- and post-operative treatment. Additionally the outcome of surgery for pancreatic cancer remains poor, largely due to late clinical presentation of most cases, and early detection, particularly in high risk groups, such as those with chronic pancreatitis, Intraductal papillary mucinous cystadenoma (IPMN) would enable earlier surgical intervention and improved survival. Use of gene expression profiles represents an innovative approach to cancer classification and prognostication and has been applied to an increasingly wide range of cancers. Recent studies have identified prognostically informative gene signatures for pancreatic cancer, and there is now an urgent need to develop methods for their measurement in routine clinical samples.