As thyroid disorders are common amongst the elderly, this study aims to evaluate the reference interval (RI) for thyroid stimulating hormone (TSH) in healthy adults aged 70 years and over. A proposed RI was determined from the Australian participants of the ASPirin in Reducing Events in the Elderly (ASPREE) randomised trial. Participants had nohistory of cardiovascular disease, thyroid cancer, dementia, or life-threatening illnesses. Participants prescribed with any thyroid-related medication at baseline were excluded. TSH levels were measured using a commercial chemiluminescence microparticle immunoassay. The RI was determined using the middle 95th percentile of the logarithmic transformed data of baseline TSH. Cox proportional hazard regression models were used to validate the RI by assessing disease incidence over time. A total of 10,995 participants had baseline TSH measures. Median (IQR) age was 73.9 (71.8-77.3) years. We propose a RI of 0.34-3.75 mU/L. TSH levels did not differ by age or sex. At baseline, there was no association between symptoms associated with thyroid disease and levels of TSH. Over the follow-up period of up to 11 years, no association was seen between baseline TSH levels and relevant disease outcomes for participants within the RI. From a group of initially healthy, community-dwelling adults aged >=70 years, we propose a RI of TSH to best represent euthyroidism. This concentration was not associated with an increased risk of thyroid related symptoms or outcomes, confirming its appropriateness for clinical use.