This paper aims at providing insights about the role played by absolute and relative income on individual well-being. Micro-data available in the British Household Panel Survey for 1996-2007 are employed for the purpose, singling out 108 reference groups on the basis of age-cohort and region. A heteroskedastic pooled panel ordered probit model with unobserved individual-speciÖc e§ects is applied, leading to Önd that comparison income exerts a negative e§ect on individual happiness, while absolute income has nearly zero ináuence. The inclusion of a measure of deprivation relative to mean income reveals that the reference group mean income e§ect is (negatively) stronger for the poor.