The WISDOM Study, that includes 45,000 women in the US, is the first large-scale study of a risk-based approach to breast cancer screening. Its goal is to determine if risk-based screening is as safe, less morbid and preferred by women. The ‘WISDOM 1.0’ cohort, enrolling since 2016, began with a risk model that integrates clinical risk factors, race/ethnicity, breast density, polygenic risk score (PRS), and sequencing for moderate- and high-penetrance germline mutations. However, in the past 7 years there have been significant advances in breast cancer risk assessment, such that we have models that, for the first time, allow us to predict the type of cancer a woman is likely to develop. Breast cancer is not one disease and women have different risk factors. Thus, it is the hypothesis of this P01 that tailoring screening and prevention recommendations to an individual’s risk as well as the type of cancer for which she is at risk, can improve the efficacy and efficiency of breast cancer screening, improve the healthcare value of screening and ultimately reduce breast cancer mortality, incidence, and screening burden. The four Projects proposed represent a comprehensive research program to advance the science and evaluation of subtype-specific risk-based breast cancer screening and prevention. In Project 1, we extend enrollment of the WISDOM Study for 5 years, applying a subtype-specific risk assessment and screening that delineates risk for fast-growing and slow-growing cancers, and assigns commensurate screening and prevention recommendations. An additional 50,000 women will be enrolled via an expanded site network. While Project 1 evaluates our initial fast vs. slow-growing risk models, Projects 2 and 3 will work to improve subtype risk assessment in two ways, utilizing 5 large, diverse study cohorts with >200,000 women and >14,000 breast cancer cases, with available imaging and germline data. Project 2 aims to develop improved PRS for subtype-specific breast cancer risk, building on our preliminary predictor of fast-growing cancer. Project 3 will apply deep-learning artificial intelligence models for subtype-specific risk based on 2D and 3D mammographic imaging, and integrate with PRS from Project 2. Project 4 will take the best integrated subtypespecific risk models and associated screening strategies from Project 3 and determine their impact on the efficiency and efficacy of screening and prevention of slow growing cancers. We will work in collaboration with the well-established MISCAN/CISNET modeling team to determine the potential impact for individuals as well as the population of women in the US. The goal is to find the optimal risk classification schema, based on reduction in cancer death as well as screening’s potential harms and improvement in overall healthcare value. Our long-term goal is to iteratively reduce breast cancer mortality, while demonstrating that the WISDOM study can ultimately serve as a continuous quality improvement platform for breast cancer screening and prevention. Four cores support this extraordinary effort to bring smarter screening to women in the US.