The thyroid gland is one of the most sensitive organs for radiation-induced oncogenesis and the magnitude of the risk from external radiation is well understood. This is not the case for internal radiation derived from the radioiodines, a matter of practical importance because of medical use and potential accidental exposure. This article reviews current knowledge derived from the follow-up of patients receiving diagnostic or therapeutic 131I and populations exposed to radioactive fallout. The latter includes the nuclear power station accident at Chernobyl and the results of atomic bomb development and testing at Hanford, the Nevada Test Site and the Marshall Islands. The most cogent information comes from Chernobyl where an epidemic of childhood thyroid cancer has followed exposure to radioiodine that was mainly 131I. Although much has been learned from this experience about the nature of radioiodine induced thyroid cancer in young children, the reconstruction of thyroid radiation doses is too preliminary to provide accurate knowledge of the risk in comparison to that from external radiation. In the Marshall Islands, much of the exposure was from short-lived radioiodines as well as external radiation, obviating the possibility to determine the risk from 131I. Exposure to 131I in the continental United States from atomic bomb testing is expected to have caused some thyroid cancers, but only in the immediate vicinity of the Nevada Test Site has any evidence of radiation-induced thyroid neoplasms been adduced. This evidence is minimally significant statistically, and not significant for thyroid cancer per se. Medical use of radioiodine has not been observed to cause thyroid cancer but very few of the patients studied were young children, the group most sensitive to thyroid radiation. Despite these limitations, this information is sufficient to make some suggestions concerning protective measures in the case of nuclear accidents and the follow up of individuals who have been exposed.