OBJECTIVES: The use of Discrete Choice Experiments to value generic preference based-measures (PBMs) such as EQ-5D-5L and SF-6D has increased and the empirical approaches used are diverse.  The aim of this paper was to review the literature using DCE to value PBMs to establish what work has been done to date, and the divergence in the approaches taken. We also aimed to establish the areas where further work is required, and to outline a future research agenda for the use of DCE to value health. METHODS: Published literature using DCE to generate values for PBMs was identified using PubMed. To assess the different DCE methods used and the similarities and differences in the approaches used, a wide range of information was extracted, including general study information, study design and methods, experimental design, modelling strategy, and results. RESULTS: Twenty published papers using 15 primary datasets with general population, student and convenience samples were identified. There is wide divergence in the study design methods used including task presentation, sample and design size, state selection procedure and presentation methods.  However the results of the papers demonstrate the feasibility of DCE to generate values for the EQ-5D-3L and EQ-5D-5L as most generally report ordered coefficient models, with low levels of inconsistency. CONCLUSIONS: The review suggests that DCE is a valid method to use to generate health state values. There is divergence in the methods leading to divergence in model characteristics.  However, one approach cannot be recommended given that there is no consensus on the best method.  Further work to test these is needed including the comparison of different design strategies and task presentation methods and a more detailed assessment of interactions between health state dimensions.  There is mixed reporting, and a checklist for papers using DCE to value health states could be developed.