The incidence of thyroid cancer has been increasing in many countries, mainly due to overdiagnosis. Given these rapid changes, we aim to assess the specific features of the thyroid cancer diagnosis epidemic and provide estimates of overdiagnosis across countries spanning five continents and identify areas in which coping strategies are needed. Two types of data were retrieved from the International Agency for Research on Cancer (IARC) Global Cancer Observatory database. The long-term annual incidence of thyroid cancer by sex and 5-year age group for all ages from 1980 to 2017 was obtained from continuous population-based registries available in the Cancer Incidence in Five Continents (CI5) plus, with 97 registries from 43 countries selected. Sex-age-specific thyroid cancer cases, overall and by subtype, along with corresponding population counts were retrieved from all registries included in the latest volume of CI5 (CI5-XII), with 385 registries in 63 countries. Annual mortality data from 1980 to 2022 were obtained from WHO, with population counts supplemented by UN population estimates. We estimated age-standardised rates (ASRs) of thyroid cancer incidence and mortality for all ages by sex using direct age standardisation, with the world population as a reference. Long-term annual trends of ASRs were compared between incidence and mortality since 1980. Subtype distribution was calculated for thyroid cancer incidence during 2013-17. We estimated the numbers of thyroid cancer cases and overdiagnosed cases and extrapolated to the whole country using a previously developed and validated method. Thyroid cancer incidence rates rose during 1980-2017 for most countries, with the highest rates seen in South Korea, Cyprus, Ecuador, China, and Türkiye for females and males. An upward trend was seen until the early 2010s, followed by a downward trend in South Korea, the USA, Canada, and Israel and some Western European countries, such as France, Italy, Austria, and Ireland. The difference between the highest and lowest incidence rates ranged from less than 10·0 per 100 000 females in the early 1980s to 101·4 per 100 000 females in 2012. For males, the difference between the highest and lowest incidence rates ranged from 2·7 per 100 000 to 23·5 per 100 000 over the study period. Mortality rates were substantially lower, with a difference between the highest and lowest rates across countries of around 1·0-2·0 per 100 000 for both sexes throughout the study period. During 2013-17, papillary thyroid cancer contributed to the large variation in ASRs of thyroid cancer incidence. The mortality rates of thyroid cancer increased with age for all countries, whereas the observed age-specific incidence rates showed an inverted U-shape in most countries. The magnitude of overdiagnosis varied across countries, ranging from no overdiagnosis (in Uganda, Zimbabwe, and Trinidad and Tobago) to more than 85·0% of thyroid cancer cases being overdiagnosed in females (in Cyprus, China, South Korea, and Türkiye). Overall, 1 736 133 (75·6%) of 2 297 057 cases were attributable to overdiagnosis, including 1 368 181 females and 367 952 males. Although the incidence of thyroid cancer has reached a plateau or decrease in some high-income countries, the magnitude of overdiagnosis is still very large and the expansion of overdiagnosis of thyroid cancer to the transitioning countries has been rapid, which makes it a global public health challenge that needs to be addressed. National Natural Science Foundation of China, Guangdong Basic and Applied Basic Research Foundation, Young Talents Program of Sun Yat-sen University Cancer Center, Italian Association for Cancer Research, and the Italian Ministry of Health (Ricerca Corrente).