The relationship between radiation exposure and thyroid cancer is well known, but whether all irradiated patients should have thyroid ultrasounds is unresolved. We have performed follow-up ultrasound examinations of patients in a cohort who were exposed to conventional external radiation during 1939-63 for benign conditions of the head and neck area prior to their 16th birthday. Of 54 subjects who had normal radionuclide scans in 1974-76 and were reexamined in 1996-97 by thyroid ultrasonography, 42 remained eligible and 34 agreed to participate in the present ultrasound study. After an additional 4-8 years of follow-up and using an ultrasound machine with increased resolution, we found 160 nodules (in 33 of these 34 subjects), compared with 96 nodules (in 29 of the 34 subjects) detected in the previous examination. Only four of the new nodules were > or =10 mm. Of the previously diagnosed large (> or =10 mm) nodules, four nodules in four subjects resolved; nine nodules in six subjects regressed to <10 mm; 14 nodules in 13 subjects remained at > or =10 mm. The four new large nodules appeared in four subjects, and six small nodules increased to > or =10 mm in six other subjects. The total volume of the thyroid nodules decreased in the 13 subjects on thyroid hormone (by 0.20 cm(3)) and increased in the 21 subjects who were not (by 0.34 cm(3), p < 0.05 by unpaired t-test). In summary, thyroid nodules are extremely common in irradiated subjects. Many new ones may be observed over time, but most are small and seen because of the increased resolution of ultrasound machines. Compared to patients on no medication, nodules in patients on thyroid hormone tended to regress. Since FNA of all thyroid nodules in irradiated patients is not feasible, ultrasound is useful in identifying those lesions that are growing.