Although the total number of patients in these various high-risk groups is relatively small, they nevertheless provide excellent models for studying the cause, natural history, pathogenesis, and treatment of pancreatic cancer. These patients would also benefit greatly from procedures capable of detecting cancer at an early stage. This knowledge would be useful for the much commoner sporadic form of pancreatic cancer, in which diagnosis is almost always late and prognosis fatal. With early diagnosis, surgical resection before the cancer's extension beyond the organ's anatomic confines could be curative. The establishment of a National Familial Pancreatic Cancer Registry is essential and would increase the availability of these invaluable families for medical research.