BACKGROUND: The ICEpop CAPability measure for Older people (ICECAP-O) and the Adult Social Care Outcomes Toolkit (ASCOT) are preference-based measures for assessing quality of life (QOL) from a broader perspective than do traditional health-related QOL measures such as the EuroQol five-dimensional questionnaire (EQ-5D). Measurement properties of these instruments have not yet been directly compared. OBJECTIVE: The purpose of this study was to compare the test-retest reliability, construct validity, and responsiveness of the three-level EQ-5D (EQ-5D-3L), ICECAP-O, and ASCOT in frail older adults living at home. METHODS: Cross-sectional data and longitudinal data were used. Parameters for reliability (the intraclass correlation coefficient) and agreement (standard error of measurement) were used to assess test-retest reliability after 1 week. We formulated hypotheses about correlations with other measures and tested these to assess construct validity and responsiveness (longitudinal validity). RESULTS: The reliability parameters for all three scales were considered good (intraclass correlation coefficient values above 0.70). Standard error of measurement values were less than 10% of the scale. Hypotheses regarding construct validity were in general accepted; the EQ-5D-3L was more strongly associated with physical limitations than were ICECAP-O and ASCOT and less strongly with instruments measuring aspects beyond health. Longitudinally, as hypothesized, mental health was most strongly associated with ICECAP-O, and self-perceived QOL, mastery, and client-centeredness of home care most strongly with ASCOT. CONCLUSIONS: Our findings support the adoption of ICECAP-O and ASCOT as outcome measures in economic evaluations of care interventions for older adults that have a broader aim than health-related QOL because they are at least as reliable as the EQ-5D-3L and are associated with aspects of QOL broader than health.