Pancreatic cancer is a highly fatal disease (five-year survival rate of only 6%) with virtually no current opportunities for meaningful efforts towards primary or secondary prevention. It is the fourth most common cause of cancer death for both men and women in the US. African Americans are disproportionately burdened with the highest pancreatic cancer incidence and mortality rates in the US - about 30% higher than the group with the next highest rates, non-Hispanic whites. The current lack of knowledge on factors that could be useful for primary prevention or for identifying high-risk groups for potential screening or intervention is a major impediment to improving the poor outlook for this cancer. Evidence has been mounting, including pilot results generated for the current application, that commensal and pathogenic oral bacteria may play a role in pancreatic cancer development. If this is the case, and if the relevant characteristics of the oral microbiota can be identified, this could be amajor breakthrough in our ability to prevent pancreatic cancer, because the oral ecology is modifiable. We see the emergence of new microbial gene sequencing and bioinformatics methods as a fundamental opportunity to study how oral microbial communities may contribute to pancreatic cancer risk and disparities. Leveraging the existing resources of two prospective, epidemiologic studies, the Southern Community Cohort Study and the Black Women's Health Study, we propose to substantially advance this emerging line of enquiry by conducting a comprehensive investigation of the oral microbiome and pancreatic cancer. We will do so within established parent cohorts that are uniquely situated to include African Americans (as well as European Americans) and to address potential racial differences in the oral microbiome as drivers of the excess burden of pancreatic cancer among African Americans. In a nested case-control study of 125 incident pancreatic cancer cases and 450 controls, using pre-diagnostic oral rinse samples, we will employ high-output genomic shotgun sequencing to characterize and compare the oral microbiome of cases and controls, hypothesizing that microbial abundance, diversity, and/or function will differ between these groups (Aim 1). Among the controls, we will also evaluate racial differences in the oral microbiome and identify any differences that may explain the increased incidence of pancreatic cancer among African Americans (Aim 2). We will further determine the association between recognized, modifiable risk factors for pancreatic cancer (cigarette smoking, obesity, red meat and processed meat consumption, alcohol consumption, type 2 diabetes) and oral microbiome attributes, and evaluate whether these exposures are associated with high-risk oral microbiome profiles (Aim 3). This study could have both immediate and long-term major impact by opening the doors for new mechanistic research as well as translational activities including manipulating the microbiome for cancer prevention, developing therapies related to the microbiome, and ultimately reducing the incidence of, and disparities in, pancreatic cancer.