As has been pointed out, more than half of the thyroidal exposure in the Marshallese was due to the more energetic and destructive short-lived isotopes of iodine. Hence the use of risk data for 31! exposure will give considerably lower estimates for hypothyroidism than would be expected for the Marshallese subjects. Several other explanations may contribute to the discrepancy between the observed and predicted number of hypothyroid patients. The estimates of thyroidal exposure, particulary that due to the short-lived isotopes of iodine, is only approximate for the reasons already discussed and could have been underestimated. Second, the hypothyroidism from which the risk estimates were derived was based largely on clinical evidence of hypothyroidism, whereas the present study has employed sensitive biochemical techniques not generally used in previous studies. A third consideration is that the early and severe thyroid dysfunction occurred in individuals exposed as very young children whereas the risk estimates are based almost exclusively on data obtained in adults with hyperthyroidism. The radiosensitivity of the young thyroid, at least for neoplastic changes, is greater than that in older patients (193). It is apparent that the true prevalence of hypothyroidism is not definable in this population because of thyroid surgery in many of the patients at highest risk. From the results in Table 8 one must consider the possibility that an elevation in TSH could have contributed to the development of the early thyroid nodularity. Such a possibility justifies the prophylactic administration of T, in the exposed Rongelap and Ailingnae population even in the absence of clinical symptoms or signs of thyroid dysfunction. This conclusion is also supported by the well-recognized relationship between irradiation, elevated TSH, and thyroid carcinoma found in animal studies (142,144) though no such relationship has been reported in man. The latent period between exposure and development of thyroid nodules appears to be longer in the groups receiving lower thyroid doses (see Figure 3). In view of the small number of cases, however, this possible association is only suggestive. The later development of thyroid hypofunction in those Rongelap people receiving lower doses would seem to be a reasonable expecta- tion because, with less cell damage and with failure of cellular replacement less critical, it would take longer for accumulative cell deficit (loss at mitosis) to result in recognizable hypofunction. The data of Table 9 are contrary to expectations in several respects. The risk of thyroid cancer is lower in both Rongelap and Utirik in the younger age groups (<l10 years at irradiation) than in the older populations. This is quite surprising as most evidence suggests that not only thyroid carcinoma but also leukemia and carcinogenesis generally are more likely induced in children than adults (99). De Lawter and Winship (150), for example, in a long-term follow-up of adults with Graves' disease who were treated with x rays, found no thyroid cancers, whereas at a rate of 3 per 106 person-years per rad, some 33 cases would have been expected. The other peculiarity is the high rate of carcinoma in the older groups (risk of 5.5 thyroid cancers/10® person-years/ rad in the Rongelap group). Finally, the induction of myedema with a thyroid dose of 1150 rads in two children is surely unexpected, as is the overall rate of hypothyroidism of 9% in the total Rongelap group. Since all of the numbers involved are small, the absolute risk figures have a high probable error. Nonetheless, it is possible to explain these data if we assume the estimates - - 80 -