PROJECT SUMMARY Patients’ health expectations lie at the foundation of their preference-sensitive choices, such as whether to pursue comfort-oriented therapies as a primary or concurrent strategy. To achieve high-quality shared decision making, patients with serious illnesses must form accurate expectations for their future health. However, many patients have inaccurate expectations, limiting their ability to make decisions concordant with their values. Chronic obstructive pulmonary disease (COPD) is an incurable lung disease of older adults and a leading cause of death worldwide. Patients with COPD and their caregivers frequently experience a high burden of symptoms, mood disorders, and difficulty coping, yet underuse advance care planning and palliative treatments. Existing evidence reveals that patients with COPD are at risk for expectation inaccuracies and our preliminary work reveals that overly optimistic expectations are associated with worse quality of life over time. Our overarching objective is to promote the well-being of patients living with COPD through the delivery of care concordant with individual patients’ values. This work proposes a novel application of behavioral theories of decision making and innovative methodologies to the common problem of goal-discordant care. This is a prospective cohort study among 420 patients with severe COPD plus their family caregivers recruited from Wake Forest Baptist Health, Geisinger Health System, and the University of Pennsylvania Health System. These three health systems represent diverse communities, including rural and Black communities where COPD is disproportionately burdensome, and have robust research infrastructures. The specific aims of this study are to: 1) identify patient and caregiver characteristics associated with inaccurate health expectations, 2) quantify associations between patients’ and caregivers’ expectation accuracy and well-being, and 3) identify mechanisms through which clinician communication influences expectation accuracy. The study team has the requisite content expertise, methodological expertise, and prior success conducting investigations of this type. Completion of this project will provide mechanistic insights into ways to promote optimal shared decision making in serious illness using state-of-the-art methods to test a novel conceptual model. This work will result in key targets for intervention development to improve expectation management and decision making among patients with serious illness.