OBJECTIVES: Self-rated health (SRH) has been shown to be a good predictor of mortality. However, there are mixed findings of the association between mortality and SRH measures with different reference points (i.e., with respect to either global or peer age group). This study assessed whether SRH measures with different reference frames influence the association of the SRH and mortality in old population.  METHODS: We analyzed data from 2000-2005 Medical Expenditure Panel Survey (MEPS) respondents in Panel 5-7, aged 60 or over, linked to the National Death Index (NDI) through 2006. To test whether the SRH measures with different reference points were comparable measures to predict mortality, two SRH measures (global and age-comparative SRHs) were applied separately and concurrently. Cox proportional hazards model was conducted, adjusted for demographic and social characteristics. RESULTS: A total of 4787 respondents were included in the analysis. More respondents were likely to assess their health as excellent or very good on the age-comparative SRH measure than on the global SRH measure (excellent, 14.7% vs. 7.6%; very good, 28.6% vs. 25.8%, respectively). In the independent models, ‘poor’ SRH ratings were the strongest predictor of mortality. Poor global SRH ratings increased mortality risk by 5.06 times and poor age-comparative SRH rating increased mortality risk by 5.02 times compared to their respective excellent ratings. When two measures were concurrently analyzed in the relation to mortality risk, both measures significantly predicted mortality and poor global SRH ratings were the strongest predictor of mortality (Hazard ratio = 2.75; 95% CI = [1.694, 4.469]).  CONCLUSIONS: In this study, both global and age-comparative SRH measures were associated with increased risk of mortality. However, we also found that the global SRH measure tended to have more predictive power than the age-comparative measure. Our findings imply that the different reference points may affect the association between SRH measures and mortality.