OBJECTIVES:  There has been a growing interest in using measures of subjective well-being (SWB) to provide broader measures of benefit in HTA than health and to enable comparisons between health care and other sectors (e.g. social care). This study compared different measures of subjective well-being (SWB) (e.g. life satisfaction, happiness) and health (EQ-5D and SF-6D) in terms of the constructs they measure and the impact of using SWB instead of health in health care assessment. METHODS:  Data came from four surveys: cross sectional surveys of inpatients (4,783) and general population (n=5709); an online survey from 5 countries (6,808); and a representative panel of (2 waves). Factor analyses identified underlying unobservable (latent) variables or domains. Sensitivity was compared using the ratio of variance explained (Eta) from the health condition variable for the SWB measures to that for the health measures. RESULTS:  Exploratory factor analyses assuming no correlation of errors suggest that the SWB measures are closely related and load on different factors from physical health. Results indicated well-being items fell into positive or negative factors with considerable cross loading. The confirmatory factor analysis and bifactor models indicate a broad well-being factor indicate a single underlying well-being factor with additional instrument or methods factors. Effect sizes for physical health conditions were much smaller for SWB measures than for health measures. There were no consistent patterns with regards to how different SWBs performed relative to health measures for the physical conditions. The multi-item SWB measures (e.g. GHQ-12, WEMWBS and ICECAP) had relatively larger effect sizes than the health measures for mental health or depression. CONCLUSIONS: Greater reliance on well-being to measure the benefits of health care would have radical implications for the priorities of health care systems.