Objectives: To examine the rates of incidence and fatality in cohorts of patients diagnosed with thyroid cancer from 1975 to 1999. Methods: This study uses National Cancer Institute's Surveillance, Epidemiology and End Results data and derives hazard functions in order to examine the fatality in thyroid cancer. Results: The study documents forms of rapidly evolving and fatal tumors as well as forms of tumor that evolve more slowly to cause death. It demonstrates that the incidences of nonfatal forms of thyroid cancer have risen dramatically in the years from 1975 to 1999-mostly due to papillary carcinomas-but that the incidences of fatal forms of thyroid cancer have remained nearly constant. Conclusions: The results of this study support the notion that many thyroid cancers are part of a reservoir of nonfatal tumors that are increasingly being overdetected and overdiagnosed.