Pancreatic cancer is the 4th leading cause of cancer death in the United States. This is in large part due to the rapidly fatal course of this disease, as the vast majority of patients die within months of diagnosis and the five- year survival rate is less than 5%. Like all cancers, pancreatic cancer is a fundamentally genetic disease caused by inherited and acquired genetic mutations. Two genome-wide association studies of pancreatic cancer, PanScan I and PanScan II, have recently been completed. These studies have identified four promising regions involved in pancreatic cancer susceptibility: ABO rs505922 (P=4.3.10-6), two correlated SNPs on chromosome 13q22.1, rs9543325 (P=3.3.10-11) and rs9564966 (P=5.9.10-8), rs3790844 (P=2.4.10-10) on chromosome 1q32.1, and rs401681 (P=3.7.10-7) on 5p15.33. The goal of this project is to conduct fine-mapping and large-scale validation genotyping of the potential pancreatic cancer susceptibility variants identified in the recently completed PanScan I and PanScan II studies, in an independent set of 4,000 cases and 4,000 controls from over 10 studies. This will be the first well-powered large-scale replication of these findings. Joint-analysis of these data with the data from PanScanI and II will also be conducted. We will also determine known risk factors for pancreatic cancer including, cigarette smoking and diabetes modify these associations.