Pre-print, Working paper: There is much discussion about using subjective well-being measures as inputs into a social welfare function, which will tell us how well societies are doing. But we have (many) more than one measure of subjective well-being. I here consider examples of the three of the main types (life satisfaction, affect, and eudaimonia) in three European surveys. These are quite strongly correlated with each other, and are correlated with explanatory variables in pretty much the same manner. I provide an overview of a recent literature which has compared how well different subjective well-being measures predict future behaviour, and address the issue of the temporality of well-being measures, and whether they should be analysed ordinally or cardinally.