As policy makers search for indices to complement existing population-based economic markers, interest has revived in ways to assess how people think and feel about their life circumstances. In this context, researchers are reconsidering the information gained from traditional global (or evaluative) measures of well-being such as life and domain satisfaction and are creating new measures to capture hedonic experience in daily life. Kahneman and colleagues (2004), for example, developed a Day Reconstruction Method (DRM) that integrates the assessment of affective experience with a detailed time diary of yesterday?s activities. The DRM provides on-line information about experience that is distinct from global reports of well-being. Our studies will address several practical and conceptual barriers to the inclusion of a DRM-like measure in large surveys of older adults. One practical barrier is the 40-75 minute length of the DRM. Pilot work on brief 10-15 minute measures that retain a limited set of DRM-features has begun in studies associated with the Health and Retirement Study (HRS). The target length of a survey measure, however, is 3-5 minutes. We currently don’t know the minimum features necessary for a short reliable and valid survey index of experienced well-being. Furthermore, we lack information about the impact of cognitive aging on the measurement of experienced well-being and we do not know if evaluative and experienced well-being have distinctive associations with health in later life. We use a multipronged approach. To begin, we compare HRS findings with DRM data collected in the Panel Study of Income Dynamics (PSID), the American Time Use Survey?s (ATUS) new well being module, and the American Life Panel?s (ALP) DRM. These secondary analyses will answer questions about: 1) the quality and comparability of responses to the fine-grained versus brief DRM measures; 2) unique correlative and predictive associations of evaluated and experienced well-being with health; and 3) differences related to life period (midlife, young old, oldest old), retirement status, and individual characteristics such as non-pathological cognitive decline and personality. We supplement these analyses with a series of focus group and experimental studies to examine the implications of age differences in episodic memory and preferences for emotional experience. Two studies will investigate measurement features expected to support or interfere with on-line (episodic) memory of recent experiences. We will compare the effects of instructions to reconstruct yesterday by time versus activities and reactivity to repeated questioning about feelings. Other studies will examine age differences in the recall accuracy of activities and time from one day to the next and memory for hedonic and non-hedonic feelings. We will also collect middle-aged and older adults’ ratings of the valence, effort, and affect associated with daily activities. Our overall objective is to establish criteria for parsimonious-but-comprehensive survey measures of subjective well-being. The information obtained will advance assessment of well-being in large surveys of older adults and our understanding of life quality in late life.