The presence of thyroid peroxidase antibodies (TPOAbs) may be considered as an indicator of adverse health outcomes. We aimed to investigate the potential determinants of TPOAb levels and to analyze the association between TPOAb titers and the risk of all- and specific-cause mortality. Baseline and longitudinal data of 13 187 participants from the ELSA-Brasil Study were analyzed. We investigated the association of TPOAb, detectability, positivity, and persistent positivity with sociodemographic and lifestyle factors using logistic regressions. Cox proportional hazards and Fine-Gray subdistribution hazard regression analyses were used to verify the association of TPOAbs with mortality. The determinants of TPOAb detectability and positivity were younger age, higher body mass index, female sex, and former and current smoking status. Black, mixed, and other self-reported races, intermediate and higher education, and heavy drinking were determinants of detectable and positive TPOAb levels. Female sex, White race, and former smoking were determinants of persistent TPOAb positivity at 2 visits, although only the female sex maintained its association at 3 visits. Moreover, after multivariate adjustment, there were associations between higher levels of TPOAbs and higher risk of cancer-related mortality among men, and TPOAb detectability and mortality by other causes among women. Sociodemographic and lifestyle-related factors were determinants of multiple TPOAb categories. TPOAb levels were associated with mortality risk; however, the low mortality rate in this sample might have compromised this finding. We suggest further studies to explore the clinical importance of detectable TPOAb levels, not only its positivity, as a potential marker of inflammation.