ABSTRACT There is a critical gap in understanding hospital- and regional causes leading to maternal mortality and severe maternal morbidity, particularly with regard to preventable deaths. Understanding and addressing these root causes of maternal mortality and morbidity is particularly urgent, especially given high rates in the US, compared to other developed countries. In addition, Black pregnant and postpartum women are 3-4 times more likely to die from pregnancy-related causes and have a two-fold higher risk of severe maternal morbidity, compared to their White counterparts. Our central hypothesis is that non-medical determinants of health, including hospital and community factors, significantly impact maternal and pregnancy outcomes. Aim 1 will link data from the large patient catchment areas of two university hospital systems (University of Missouri and University of Utah) with geocoding, social security death files, obituary files, and PCORnet to elucidate the impact of non-medical determinants of health, on rates of maternal mortality and morbidity, and to assess the extent to which these factors explain or predict disparities among Black women. We hypothesize that non-medical determinants of health will have a significant association with mortality and morbidity, and will help explain these disparities between Black (versus White) pregnant and postpartum women. Aim 2 will interrogate de-identified healthcare records from the Cerner Corporation’s multi-institutional Oracle Real-World DataTM system to identify hospital-level factors associated with maternal mortality and morbidity. We hypothesize that hospital-level and regional factors such as uneven distribution of medical services (including rural status), poverty, maternal levels of care, and patient demographics and comorbidities will have significant impact. The proposed research is innovative, as it will (a) use data to develop and validate a prognostic scoring tool for maternal mortality and morbidity; (b) assess hospital-level factors that may contribute to maternal mortality, using data from 128 separate health systems in Cerner’s database; (c) integrate geocoding and linked death data to enable better estimates of non-medical determinants of health factors that contribute to maternal mortality and morbidity; and (d) facilitate a rapid “scale-up” to the national level, given the multi-site nature of the Greater Plains Collaborative (GPC)/PCORnet data system.