Abstract Prospective cohort study is instrumental in assessing cancer risk factors, identifying biomarkers of cancer risk and early detection, and constructing integrative risk prediction model. However, developing a large prospective cohort with longitudinal, repeated questionnaire data and biospecimen collection is often cost-prohibitive. Building upon a large-scale health screening program in Taiwan with existing questionnaire data, medical test results, and biospecimens, we have developed a large prospective health screening cohort – MJ Cohort. MJ Cohort consists of individuals undergoing routine health screening conducted by the MJ Group, the largest health management institution in Asia. Over 600,000 unique participants have been enrolled with detailed baseline questionnaire data collected on demographics, lifestyle factors, occupation, medical history, family history of chronic diseases and cancer, and dietary pattern. Participants undergo a series of medical tests (103 items) for blood, urine, anthropometric measurements, functional tests, and physical examination. Approximately half of all the participants have longitudinal questionnaire data and medical test results. A substantial number of participants also had blood samples collected and stored. We have utilized this unique cohort for a number of research initiatives. For example, we studied the mortality reduction benefits from leisure-time physical activity and found that an average 15 minutes per day of moderate intensity exercise reduced all-cause mortality by 14% and all-cancer mortality by 10%, and gained 3 years in life. Every additional five minutes of daily exercise is predicted to further reduce 2% of all-cause and 1% of all-cancer mortality. We constructed a series of liver cancer risk prediction models, with the highest prediction accuracy, measured as the AUC (area under the ROC curve), reaching 0.933 when combining two simple serum markers of liver function (ALT and AST) with hepatitis infection. However, individuals could substantially reduce their risk by improving their lifestyles, stopping smoking, reducing drinking, becoming physically active, or managing diabetes. We are also using this unique cohort to identify patients at high risk of progression among those with pre-cancerous conditions and to identify biomarkers of aging and dissect the complex relationship of aging-related chronic diseases and cancer. The extensive questionnaire data, medical tests, and existence of strong risk factors and biomarkers will allow us to construct powerful risk prediction models for selected cancers, which will provide evidence-based recommendations for lifestyle and behavioral changes to achieve reduction in cancer burden and identify high-risk individuals and minimally invasive early detection markers for common cancers. The unique and highly efficient infrastructure for participant enrollment, data collection, medical tests, and follow up within the framework of a world-leading health screening center make this cohort extremely cost-efficient. This unique health screening cohort therefore provides an invaluable new avenue for risk prediction, prevention, and early detection research. Citation Format: Xifeng Wu. Health screening cohort: A new avenue for risk prediction, prevention, and early detection research. [abstract]. In: Proceedings of the Eleventh Annual AACR International Conference on Frontiers in Cancer Prevention Research; 2012 Oct 16-19; Anaheim, CA. Philadelphia (PA): AACR; Cancer Prev Res 2012;5(11 Suppl):Abstract nr CN02-01.