BACKGROUND: Selenium is an essential trace element with antioxidant property. Decreased serum selenium concentration with aging had been found in previous report. In this study, we aim to investigate the association between serum selenium and the inflammatory cytokine interleukin-6 in the elderly living in long-term care facilities in Taiwan. MATERIALS AND METHODS: A total of 336 subjects aged 65 years and older (range of age: 65–101 years) were recruited from eight long-term care facilities in 2002–2003. Baseline characteristics, anthropometric indices, and biochemical data were obtained. Selenium deficiency was defined as serum selenium concentration < 80 μg/L. Multiple logistic and linear regression analyses were used to examine the relationships between selenium deficiency and interleukin-6 (divided into quartiles). RESULTS: The prevalence of selenium deficiency was 35.6% in men and 43.2% in women, respectively. After adjusting for potential confounders using multiple logistic regression analysis, interleukin-6 quartiles were significantly associated with selenium deficiency. Compared to the interleukin-6 quartile I, the adjusted odds ratios of having selenium deficiency for interleukin-6 quartile II, III, IV were 1.00(0.50∼2.01), 1.24 (0.62∼2.50), and 2.35(1.15‖4.83), respectively. The increasing odds ratios for selenium deficiency in higher interleukin-6 quartiles revealed dose-response effects (p < 0.05). Moreover, multiple linear regression analysis showed that serum selenium was significantly inversely associated with interleukin-6 after adjusting for potential confounders. CONCLUSIONS: Serum selenium was inversely associated with inflammatory cytokine interleukin-6 among elderly living in long-term care facilities in Taiwan. Monitoring serum selenium should be considered in these institutionalized elderly.