This review covers new mechanistic information spanning the past 10 years relevant to normal and abnormal thyroid growth and function that may assist in the risk assessment of chemicals inducing thyroid follicular cell neoplasia. Recent studies have shown that thyroid regulation occurs via a complex interactive network mediated through several different messenger systems. Increased thyroid-stimulating hormone (TSH) levels activate the signal transduction pathways to stimulate growth and differentiation of the follicular cell. The important role of TSH in growth as well as in function helps to explain how disruptions in the thyroid-pituitary axis may influence thyroid neoplasia in rodents. New investigations that couple mechanistic studies with information from animal cancer bioassays (e. g., sulfamethazine studies) confirm the linkage between prolonged disruption of the thyroid-pituitary axis and thyroid neoplasia. New initiation/promotion studies in rodents also support the concept that chronic stimulation of the thyroid induced by goitrogens can result in thyroid tumors. Some of these studies confirm previous suggestions regarding the importance of chemically induced thyroid peroxidase inhibition and the inhibition of 3,3',5, 5'-tetraiodothyronine (T4, thyroxine) deiodinases on disruption of the thyroid-pituitary axis leading to thyroid neoplasia. Some comparative physiologic and mechanistic data highlight certain differences between rodents and humans that could be expected to confer an increased vulnerability of rodents to chronic hypersecretion of TSH. New data from epidemiologic and molecular genetic studies in humans contribute further to an understanding of thyroid neoplasia. Acute exposure to ionizing radiation, especially in childhood, remains the only verified cause of thyroid carcinogenesis in humans. Iodine deficiency studies as a whole remain inconclusive, even though several new studies in humans examine the role of dietary iodine deficiency in thyroid cancer. Specific alterations in gene expression have been identified in human thyroid neoplasia, linked to tumor phenotype, and thus oncogene activation and tumor-suppressor gene inactivation may also be factors in the development and progression of thyroid cancer in humans. An analysis by the U.S. EPA Risk Assessment Forum, prepared as a draft report in 1988 and completed in 1997, focused on the use of a threshold for risk assessment of thyroid follicular tumors. New studies, involving several chemicals, provide further support that there will be no antithyroid activity until critical intracellular concentrations are reached. Thus, for chemically induced thyroid neoplasia linked to disruptions in the thyroid-pituitary axis, a practical threshold for thyroid cancer would be expected. More information on thyroid autoregulation, the role of oncogene mutations and growth factors, and studies directly linking persistently high TSH levels with the sequential cellular development of thyroid follicular cell neoplasia would provide further confirmation.