OBJECTIVE Studies based on Western populations showed a negative relationship between dehydroepiandrosterone sulfate (DHEAS) level and mortality, but no study examined this relationship in a non-Western country. We use data from a large, nationally representative sample (n = 963) of older Taiwanese to investigate whether serum DHEAS, predicts subsequent mortality during a 3-year period (2000 to 2003) and whether an effect remains after controlling for baseline health status. METHODS Baseline data collection included an individual interview, physical examination, and blood sample. A logit model is used to test the relationship between DHEAS level and risk for mortality, controlling for age, sex, and smoking status. RESULTS Results show a marginally significant inverse relationship between DHEAS level and 3-year mortality risk. Participants with low DHEAS levels (<54.5 microg/dL) have 64% greater odds of dying than those with higher DHEAS levels (p < 0.06). After adjusting for various indicators of health status in 2000, the odds ratio (OR) for low DHEAS level remains substantial (OR = 1.41), but not statistically significant. CONCLUSIONS Although the analysis is limited by the short follow-up and small number of deaths, results are consistent with the notion that DHEAS level has a sizeable effect on mortality.