and required upward estimates of midline depth doses (by about 50%) when compared with the same air doses monitoring bilateral narrow-beam radiotherapeutic exposures. 8. Possible effects of location, clothing, shielding, bathing in the lagoon, etc., could not be estimated for each individual; this necessitated assignment of a common whole-body radiation dose estimate to each population. 9. Estimates of internajly deposited radioisotopes were based on assays of pooled urine specimens obtained two weeks after the accident and on individual 24-hour urine collections obtained between one and seven weexs after exposure. Radiation doses thus are not direct measurements but are estimates derived from all sources available at the time and subsequently developed (see Dose Assessment, Appendix II). Figure 1 presents isopleths of estimated gamma deses three feet in air (neglecting sea water dilution) integrated over the 96-hour period following the Bravo detonation. Despite the relative consistency of dose estimates, the actual values derived are subject to some uncertainty, particularly for doses due to internal deposition, and these uncertainties wust be considered when evaluating potential biological effects. For example, the recent increase in incidence of thyroid neoplasms in the least exposed population (Utirik) raises additional possibilities of chemical fractionation within the rapidly rising, cooling fireball with resultant non-uniform deposition of specific radionuclides as "kot spots" within the fallout pattern, a phenomenon well documented in other nuclear weapon tests (40). For report purposes, certain dosimetric values have been assigned to is- land populations (Table 1), but individual doses within each group are obviously subject to broad variation (see Appendix II and Section IX). This is especially pertinent in regard to estimated thyroid doses, since these are a result of a number of imprecisely definable variables. This is exemplified, for example, by the appearance of growth retardation secondary to severe hypothyroidism in some children belonging to a group presumed to have thrroid doses approximating 1150 rads, well below the known level for thyrcid ablation. Table 1. Estimated radiation doses in exposed rfopulations. Estimated thyroid dose (rem) Atoll Rongelap Ailingnae Utirik Number Estimated whole-body affected*_.. gamma dose (rem) 67 19 163 _ 175 69 14 - by age at exposure <10 yr 10-18 yr > 18 yr 810-1800 275-450 60-95 334-810 190 30-60 335 135 30 *Includes in utero exposures (3 on Rongelap, 1 on Ailingnae, and 6 on Utirik>.