The purposes of this article are to use replicated secondary data analysis to summarize information about the relationship between health and subjective well-being and to assess the strengths and weaknesses of replicated secondary data analysis as a mode of research synthesis. The findings from thirty-seven replications in seven surveys suggest a moderate and robust relationship between self-rated health and subjective well-being. Physician-assessed health, in contrast, exhibits weaker and less robust associations with subjective well-being. Further, the relationship between health and subjective well-being is conditioned by age and is stronger for measures of negative than positive affect. The principal advantages of replicated secondary data analysis, vis-a-vis other modes of research synthesis, are cost-effectiveness, increased ability to apply multivariate statistical techniques, and greater control and flexibility for the investigator. We suggest, nonetheless, that different modes of research synthesis can best be used for different purposes.