The UK tariff for quality of life associated with the health states in the EQ-5D is probably not consistent with the preferences of Swedish health economists. This is worrying in view of the widespread use of the tariff values and the fact that health economists likely are better able than ordinary citizens to report their preferences for health states in a valid and reliable manner. We suggest this result is taken into account when the EQ-5D instrument is used, and that researchers should be cautious in using the UK (or any other) value sets. Our results also indicate that the variation across citizens in preferences for health may be a more complex issue than previously observed and deserves further study. An intriguing question for the future is to what extent health economists use methods and instruments that they themselves do not believe in.