whose thyroid doses ranged from about 400 to 5200 cGy, only 1 of 28 (4%) had the diagnosis of occult papillary carcinoma. If only those Rongelap individuals with doses less than 2000 cGy are analyzed, 1 of 13 (8%) had an occult carcinoma, and for those receiving over 2000 cGy Therefore, the high” the prevalence is 0%. radiation dose received on Rongelap may have decreased rather that increased the incidence of the occult tumors. But a factor that renders any conclusion of questionable value is that in many of the surgical explorations the entire thyroid gland was not removed and subjected to the close histologic examination that has been used in most studies on occult papillary carcinoma prevalence. The Marshallese data may, therefore, underestimate the prevalence of these lesions, particularly in the multinodular adenomatousgoiters of the exposed Rongelap children. The combined number of occult papillary carcinomas and overt carcinomasin the Rongelap and Utirik groupsis virtually identical, being 7.0% in the former and 6.6% in thelatter: Rongelap Exposed persons Occult CA Overt CA 86 1 5 Total CA Exposed/Total CA 6 7.0% Utirik 167 6 5 11 6.6% There are proportionally more carcinomas in the Rongelap group and more occult papillary carcinomas in the Utirik. One might wonder whether radiation exposure had the effect of inducing or hastening a change toward increasing virulence in the usually benign-acting “occult” lesions. However, such an interpretation does not take into account either the limitations on technique of histological examination of the thyroid mentioned in the preceding paragraph or the extensive thyroid injury in those in the Rongelap group who received more than 2000 cGy to the thyroid. 8) Did thyroxine suppression decrease the incidence of benign and/or malignant nodules in radiationexposed persons? Administration of thyroxin for the purpose of suppression of development of thyroid nodules in Marshallese who had been living on Rongelap at the time of exposure was initiated in 1965 shortly after the first thyroid nodules were detected (Conardet al., 1967). The distribution of thyroxine, subsequently extendedto include those persons who were on nearby Ailingnaeatoll, has been continued up to the present, with dose being determined by the results of yearly tests of thyroid function. Utirik patients are not routinely managed with thyroxine suppression. It is given to them only when clinically indicated as noted below. Every six months a supply of tablets is handed out to each exposed Rongelap person, whether or not that person appears for examination, and clinical decisions relating to thyroxine use are made each year by endocrinologic consultants who accompany the medical team during their work in the Marshall Islands. Thyroxine is also given to all persons who had thyroid surgery under the auspices of the Marshall Islands Medical Program, whether exposed or not, for replacement and suppression. The value of suppressive therapy in prevention of thyroid cancer and benign nodule formation is not clearly determined. Various studies, for the most part carried out on persons who previously had treatmentfor thyroid nodules, have indicated (1) no nodule suppressive effect (DeGrootet al., 1983), (2) no cancer suppressive effect (Cady etal., 1983), (3) no suppression of benign nodules (Geerdsen and Frolund, 1984), and (4) a decrease in benign nodules but not malignant nodules (Fogelfeld et al., 1989). One study found that thyroxine reduced the numberof recurrences in those who had previously undergone therapy for papillary thyroid carcinoma (Schneider et al., 1986), although the number of patients not given suppressive therapy was small. The timing of thyroxine prophylaxis may be an important factor in determiningits effectiveness; if started some years after exposure its value may be lessened (DeGrootet al., 1983). Any conclusions derived from the results of the Marshallese program have scientific limitations. The Rongelap group has been receiving thyroxine suppression since 1965 but it is known that compliance with this regimen is poor, estimated at no better than 50% (Adamset al., 1983). The suppression was not initiated until 10 years after 23