Objectives: Unlike cost benefit analysis, cost effectiveness analysis (CEA) is restricted to length of life and health-related quality of life as integrated measures of benefit. Valuation is based on individual preferences for health states. Yet dimensions of social value may exceed those driven by (aggregated) individual preference satisfaction. Multi-criteria decision analysis (MCDA) and social cost value analysis (SCVA) were proposed as alternatives to the conventional logic. Both approaches need support by robust evidence on social preferences. Methods: A literature review identified empirical studies reporting health care resource allocation priorities. On this basis, the authors deliberated implications and ways forward. Results: The review revealed social preferences for health care resource allocation including, beyond economic efficiency, (a) preferences primarily related to the health state (severity of initial health state and urgency of an intervention); (b) social preferences related to patient attributes (such as younger age, parent and caregiver status, and non- smoker); (c) social preferences with regard to allocation rules and a dislike against all-or-nothing allocation decisions and against discrimination of certain patient groups – apparently related to rights-based reasoning, including a relatively minor role of treatment costs per patient. Limitations of the literature were identified to include heterogeneity of study designs, small size of many studies, and potential bias due to framing effects and unstable preferences in some surveys. Conclusions: The expert group agreed that a European Social Preference Measurement (ESPM) study should address the limitations above. The study will be conducted in two phases (pilot study in Switzerland before international roll-out), adhere to a discrete choice experiment design, and address how the public valuates key attributes of health care interventions, explore international similarities and differences with respect to the weighting of the attributes and their interaction, and assess robustness to framing effects. Design of the study will be presented for discussion.