This dissertation compiles three papers that focus on challenges in research on and practice of palliative care. The first chapter reports on a survey of 61 leading academic researchers in palliative care, who were asked their opinions on research priorities and barriers to better research in the field. Qualitative methods were used to extract the seven leading research recommendations and four barriers to better research generated by the purposively interdisciplinary group. The next two chapters use the Wisconsin Longitudinal Study to test hypotheses about the correlates and triggers of advance care planning (ACP) in older adults, cross-sectionally and, for non-planners, at a follow-up seven years later. The results presented in Chapter 2 identify several dimensions of health that are associated with formal and informal ACP cross-sectionally. Over time, changes in health have little affect on the likelihood of a non-planner completing ACP by Time 2, seven years later. In Chapter 3, I applied theory of social diffusion to ACP, hypothesizing that the planning status of a spouse or sibling would increase a non-planner's likelihood of doing planning by Time 2. The cross-sectional analysis found strong associations between spouse only and spouse and sibling status on formal and informal ACP of the focal individual, respectively. Over time, however, having a spouse or sibling with formal (but not informal) ACP affected the focal individual's likelihood of going on to do planning, but having both a spouse and sibling with formal ACP at Time 1 had an attenuating effect on the focal individual's likelihood of planning. In the final chapter I draw connections between the three research activities and suggest directions for future research. Most pressing is the need for evaluation of existing education programs to support older adults as they complete ACP and the development of novel programs targeting individuals who are least likely to engage in ACP. The growing body of academic research on factors that facilitate ACP should be translated into more effective supports for conversations about end-of-life health care.