iodoprotein levels, leading to a false interpretation of protein-bound iodine (PBI) determinations. It became apparent only then that low serum thyroxine (T,) levels in some of the children had probably been masked by high levels of iodoprotein. Several children with slight growth retardation had lowered Ty, levels. The hypothyroidism appeared to be from primary thyroid damage and not secondary to pituitary damage, since tests for growth hormone in several growth-retarded children were normal, and their serum thyroid-stimulating hormone (TSH) levels were elevated. In 1963, 9 years after exposure, a 12-year-old exposed Rongelap girl was found to have an asymptomatic nodule of the thyroid gland. Development of thyroid nodules has continued in other exposed Rongelap and Ailingnae people and in the past decade in the Utirik group. Two of three Rongelap children exposed in utero developed thyroid nodules during the past 6 years. At this time (1981) a total of 46 exposed Marshallese (29 in the Rongelap/Ailingnae group and 17 in the Utirik group) have developed nodules, and 42 of these have undergone surgery. In the unexposed comparison population 35 of 600 people in the age-matched control group have been found to have nodules and 14 of these have undergone surgery (see Section IX.C. and Appendix IV, Table 1). In 1965 the seriousness of the development of thyroid lesions in the Rongelap people became more evident, and preventive treatment with thyroid hormone was instituted in the Rongelap group. Later (1969) the Ailingnae group was also included. The difficulties in carrying out the treatment regimen. will be referred to in Section IX.C.4. With increasing development of thyroid abnormalities it has become apparent that the original control group of unexposed Rongelap people was numeri- cally inadequate for proper evaluation of the thyroid findings in the exposed people. Therefore, during the past decade thyroid surveys have included all available unexposed Rongelap and Utirik people as well as some people at Likiep and Wotje atolls. More than 1200 of these people have been examined. About half of them were in the same age range as the exposed people, and the others were in the younger age group (born after 1954). During the last seven years, specific radioimmunoassay techniques for measurement of serum Ty, and TSH have been employed to monitor the population for evidence of hypothyroidism (1,36,110). Results have shown that at least one-half of those who had thyroid surgery, and supposedly were taking thyroid replacement medication, had significant elevations in serum TSH on one or more occasions, indicating lack of compliance with the thyroxine replacement schedule and also indicating the inadequacy of residual thyroid tissue in patients with subtotal thyroidectomies to maintain the euthyroid state without compenSsatory increases in TSH secretion. These observations led to attempts to determine the degree of thyroid function in exposed populations of both Rongelap and Utirik, which necessitated temporary discontinuation of the thyroid hormone in the Rongelap group. (The Utirik group was not routinely receiving such treatment.) CSF In earlier studies, discussed in detail in the 20-year report (1), a number of individuals with apparently normal thyroid glands were found to have subnormal increments in plasma T, 24 hours after receiving 10 units of TSH intramuscularly. The mean increment in Ty, in 25 exposed Marshallese was 2.35 ug/d1 (SD = 1.2), whereas the increment in a control group of unexposed euthyroid Rongelap and Utirik individuals was 4.2 ug/dl (SD = 1.3). This was significantly greater than in the exposed subjects (p > 0.001). A number of mA - 54 -