of radiation dose to the thyroid are too low, at least in some of the people, by a factor of 3 or 4. This would account for the high incidence of hypothyroidism. It would also account for a relatively low risk for thyroid cancer in the children, since such a level of radiation would be expected to damage the thyroid so severely that cell death would, in many instances, inhibit cell division that might result in malignant changes (195). Since the dose calculations, derived from counting of a single pooled specimen of urine, require many assumptions, an underestimate is not unlikely. Tt is quite likely that the final tally of thyroid abnormalities in the Marshallese is incomplete at this time since new lesions are still becoming clinically evident. The mean latent period for radiation-induced thyroid tumors may be as long as 30 years (196). Recent observations suggesting that as many as 25 years may pass between exposure to radioiodine and the appear- ance of even minimal biochemical thyroid abnormalities make it clear that long-term follow-up of the Marshallese must continue. - 81-