This study examined the association of the quality of built environment in the neighborhood and maternal mental health. In Aim 1 of its 3 aims, we developed predictive models to predict and subtype postpartum depression (PPD) using clinical and built environment information. The predictive model was constructed using a combination of machine learning models, and achieved an AUC of over 0.9 in prediction. The subtyping work identified 3 subgroups of women based on their risk of PPD. Women who experienced higher rates of PPD were more likely to reside in neighborhoods with homogeneous land use, lower walkability, lower air pollutant concentration, and lower accessibility to retail stores after adjusting for age, neighborhood average education level, marital status, and income inequality. In Aim 2, interviews were conducted with clinicians, including obstetrics-gynecologist, reproductive psychiatrist, and pediatricians, on their views of how to leverage information on neighborhood built environment in routine clinical care. In Aim 3, we constructed a list of variables representing social determinants of health including built environment across the states of New York, New Jersey, Connecticut, and Pennsylvania. Using these, a univariate and multivariate analysis was conducted with mothers’ PPD and children’s allergy status as the outcomes and the environmental variables as predictors. Outputs of the overall study includes journal publications (including being named as Editor’s Choice at the Journal of Affective Disorders), invited presentation at Epic Corporation, and findings that lead to ongoing preparation for a clinical trial at Weill Cornell Medicine.