Objective: This study examines and compares respondent, interviewer, and physician ratings of overall health. Method: Data are from the 2006 Social Environment and Biomarkers of Aging Study, a nationally representative survey of older adults in Taiwan. Ordered probit models are used to examine factors associated with self- and external assessments of health and discordant health ratings. Results: Our results suggest similarities and differences in factors influencing health ratings across evaluators but a high level of interevaluator disagreement in ratings. Discrepancies in ratings between physicians and both respondents and interviewers are associated with the greater weight given to functional limitations and psychological well-being in interviewer and respondent ratings and to the importance of clinical measures or risk factors of illness and mortality in physician assessments. Discussion: Interviewer and physician assessments may be complementary to self-assessed health measures. The importance and implications of these findings for future research are discussed.