Ideally, decisions in healthcare regarding patient care are based on evidence from trials and formal meta-analyses. The merits of these methods and their impact on clinical practice cannot be overestimated. However, their focus on single outcomes limits their use in multi-attribute decisions. The discrete choice experiment (DCE) technique has been introduced in healthcare to arrive at optimal decisions, dealing with multiple outcomes and preference heterogeneity among decision-makers. The technique, originating from mathematical psychology, has uses in marketing, transport and environmental economics as well, where it is used to predict individual and collective choices. In a DCE, respondents are confronted with a selected series of choices between two or more options, where options are characterized by attributes each with a given attribute level. The stated preferences, after complex Nobel prize-winning econometric modeling, result in the ranking of all possible options. Currently the lack of insight into external validity hampers DCE taking a key position in healthcare: are stated preferences consistent with actual healthcare utilization (so-called revealed (true) preferences)? To support claims based on DCEs, external validity is widely recognized as the single most important research question. My project aims to provide improvements to the DCE technique for healthcare and beyond by 1) measuring external validity; and 2) unraveling its determinants. Specifically: 1) External validity will be estimated directly at the patient level; 2) Two sources for inconsistency between DCE-outcomes and actual utilization are studied in-depth: A) Imperfections of DCE design and analysis (the role of the researcher); B) Effects of personal characteristics (the role of the respondent). I will use a mixed methods design and three areas (influenza vaccination, colorectal cancer screening, and prostate cancer treatment). The project findings will lead to dual knowledge utilization: for those who investigate multi-attribute decisions, and for those who make multi-attribute decisions in healthcare.