This paper is based on the results of a national survey of the quality of life of 999 randomly sampled people aged 65 and over, living at home in Britain. The survey was semi-structured, and a sample of survey respondents was followed up and interviewed in-depth in order to explore their perceptions of quality of life in full. Comparisons are made here between the results of (i) a hierarchical multiple regression model based on theoretically derived indicators of survey respondents' ratings of their overall quality of life, with (ii) the same respondents' own definitions of quality of life, categorised from their responses to open ended survey questions, and (iii) the views of a sub-sample of these survey respondents who were subsequently interviewed in greater depth. Respondents were asked the open ended questions on quality of life at the outset of the survey interview in order to prevent any respondent bias from the structured measures used. The core components, and the central planks, of quality of life, which were consistently emphasised by the three approaches, were psychological characteristics and outlook, health and functional status, personal and neighbourhood social capital. The lay models also emphasised the importance of financial circumstances and independence, which need to be incorporated into a definition of broader quality of life.