OBJECTIVES Health states defined by multi-attribute instruments such as the EQ-5D-5L can be valued using Time Trade Off (TTO) or Discrete Choice Experiment (DCE) methods.  A key feature of the tasks is the order in which the health state dimensions are presented to respondents.  This is because respondents may use a variety of heuristics to answer the questions (for example focusing on the first dimension presented), and therefore the order of the dimensions may impact on the importance assigned to particular health state descriptions. The objective of this study is to assess the impact of different EQ-5D-5L dimension orderings on health state valuations. METHODS Preferences for EQ-5D-5L health states were elicited from a broadly representative sample of members of the UK general public. Respondents were allocated to one of three EQ-5D-5L dimension orderings, and completed 10 TTO and 9 DCE tasks via a computer assisted personal interview carried out in their home. Differences in mean values and the relative importance of the coefficients across the arms were compared using difference testing and regression analyses. RESULTS Descriptive analysis suggests that there are minimal differences between the mean TTO health state values across the different dimension orderings.  Regression analysis suggests that the magnitude of the dimension coefficients differs across the different dimension orders (for both TTO and DCE), but there is no clear pattern. CONCLUSIONS There is some evidence that the order in which the dimensions are presented impacts on the coefficients, which may impact on the health state values provided.  The order of dimensions is an important consideration in the design of health state valuation studies.