Biliopancreatic malignancy is one of the leading causes of cancer death in the Western world. Defining at risk groups has been difficult. Diabetes mellitus and pancreatitis increase the risk of pancreatic carcinoma, and inflammatory bowel disease and associated sclerosing colangitis increase the risk of biliary tract malignancy. Pancreatic carcinoma has also been described in pedigrees with inherited cancer predisposition. Extensive molecular profiling of pancreatic carcinomas has been accomplished over the past few years, but similar knowledge in other biliopancreatic malignancies is lacking. In almost all pancreas cancers at least one alteration will occur out of a combination of K-ras mutations and inactivation of the tumor suppressor genes p16/MTS1/ink4a, p53 and DPC4/Smad4. Mutations of K-ras and p16 have been described in hyperplastic and dysplastic pancreatic ductal lesions believed to be the non-malignant precursors of pancreatic carcinoma. Detection of K-ras mutations in clinical samples (biliopancreatic secretions, stool, duodenal aspirates, and blood) identical to ones present in primary pancreatic cancers and/or their precursor ductal lesions has been reported in pilot studies. Recently detection of 18q deletions (at the DPC4 locus) in pancreatic secretions from early pancreatic cancers was also reported. These advances raise the possibility that within well defined at risk groups it will be possible to use a combined set of molecular markers to screen clinical samples and detect early pancreatic cancer or even pre-malignant lesions. The fulfillment of this promise will depend on proving the role of molecular screening in decreasing morbidity and mortality, which will require well designed clinical studies.