Article dans une revue: We here use data from the Wisconsin Longitudinal Study (WLS) to provide one of the first analyses of the distal (early-life) and proximal (later-life) correlates of older-life subjective well-being. Unusually, we have two distinct measures of the latter: happiness and eudaimonia. Even after controlling for proximal covariates, outcomes at age 18 (IQ score, parental income and parental education) remain good predictors of well-being over 50 years later. In terms of the proximal covariates, mental health and social participation are the strongest predictors of both measures of well-being in older age. However, there are notable differences in the other correlates of happiness and eudaimonia. As such, well-being policy will depend to an extent on which measure is preferred.