PURPOSE: Thyroid dysfunction is common, particularly among older women. The safety of thyroid hormone use and long-term prognosis of hyperthyroidism remain controversial. We performed a prospective cohort study to examine the relationship among thyroid hormone use, previous hyperthyroidism, abnormal thyroid function, and mortality. METHODS: We studied 9449 community-dwelling white women aged > or =65 years followed for 12 years. For analyses of thyroid function, we performed a nested case-cohort in 487 women using a third-generation thyroid-stimulating hormone assay. Causes of death were adjudicated based on death certificates and hospital records. RESULTS: Twelve percent of the 9449 women took thyroid hormone at baseline, and the mean duration of thyroid hormone use was 15.8 years; 9.4% of participants reported a history of hyperthyroidism. During 12 years of follow-up, 3159 women died (33%). In multivariate analysis, mortality among users of thyroid hormone was similar to that observed for nonusers (relative hazard [RH] 1.11, 95% confidence interval [CI], 0.98-1.24, P=.09). Previous hyperthyroidism was associated with a higher risk of all-cause mortality (RH 1.20, 95% CI, 1.06-1.36), particularly cardiovascular mortality (RH 1.46, 95% CI, 1.20-1.77). Low (< or /=0.5 mU/L) or high (>5 mU/L) thyroid-stimulating hormone levels were not associated with excess total or cause-specific mortality, but the power to detect these relationships was limited. CONCLUSIONS: Among older women, thyroid hormone use is not associated significantly with excess mortality, but previous hyperthyroidism may be associated with a small increase in all-cause and cardiovascular mortality. Additional long-term studies of hyperthyroidism and its treatment should further explore these findings.