The risk of developing thyroid cancer and other thyroid neoplasms after radiation exposure is well known, but specific modifiers of the dose-response relationship are not. We have identified 4296 subjects who received treatment before their sixteenth birthday with orthovoltage radiation for benign conditions in the head and neck area. Individual thyroid dose estimates were calculated for 3843 subjects. Of the 2634 subjects who have been found, 1043 have developed thyroid nodules of all types, and 309 have developed thyroid cancer. The radiation dose-response relationship was consistent with a linear excess relative risk model for thyroid cancer and thyroid nodules within the range of observed doses. Women developed thyroid cancer and thyroid nodules at a higher rate, but the slopes of the dose-response curves were the same for men and women. Age at radiation exposure was a significant factor of the risk, with a lower age at exposure associated with a higher risk. To determine the effect of the wide publicity and our screening program, which began in 1974, we compared the dose-response relationship for cases diagnosed before and after 1974. The overall rates increased dramatically after 1974, but the estimates of the slopes of the dose-response curves were not statistically different. The slope of the dose-response curve for thyroid neoplasms appears to have reached a maximum 25-29 yr after radiation exposure, but the dose response continued to be elevated at the end of follow-up. These data are consistent with the tumorigenic effects of radiation lasting at least 40 yr.