The lifetime risk of overt hypothyroidism is around 5%, and this disease is usually preceded by subclinical hypothyroidism, which has an even higher prevalence (estimated to be up to 9%). Hypothyroidism has been linked with cardiac dysfunction, atherosclerosis, hypertension and coagulopathy. Intuitively, this increased morbidity is expected to shorten patients' lifespan, but definitive data are lacking on whether either of these hypothyroid states (particularly overt hypothyroidism) increase mortality. Study findings are inconsistent and, overall, the pooled data do not demonstrate increased mortality in patients with either subclinical or overt hypothyroidism. However, none of the available studies was adequately designed to answer this question. This Review discusses major shortcomings in those studies, such as population dissimilarities, hypothyroid state classification and misclassification, the inclusion of nonthyroidal illness, drug interference from concurrent therapies, serious comorbidities (for example, cardiovascular disease), differences in duration of follow-up and the number of levothyroxine-treated individuals. Taken together, the data exhibit little evidence of systematic bias and no strong scientific proof of increased mortality related to either subclinical or overt hypothyroidism. Future studies, however, should take the above-mentioned shortcomings and potential genetic confounding into consideration.