Five samries were optained sequentially from 16 persons during the January 1979 field trip to derermine the variability inherent in the 24 hour urine sample program. The results of this study are listed in Table 7. For 13765, the mean biological and counting variability (one standard deviation) associated with a single urine sample is 32%. For 905, most of the results were less than the minimum detection limits of the system or the average of the 5 urine sample results had an associated standard deviation which was larger than the result. Consequently, only 6 sample results were used to determine the biological and counting variability of the 705, urine data. deviation associated with this result is 65%. The mean standard The counting error contributes 15% of the variability while other sources of variation account for 50%. These other sources are most likely related to the day to day metabolic changes norMal.y exhibited by an individual. B. Whole-Body Counting Whole-body counting measurements on the Bikini population that were conducted in 1974, 1977, 1978, 1979 and 1980 are presented. The body burden mea- surements were performed by two different organizations; consequently, the experimental design included a mechanism to ensure that previous and current resuits are directly comparable. Key detection components were duplicated and the sys- tems were calibrated in the same manner (CO 63). The operational procedures and counting geometries were basically similar, and an intercomparison study was conducted using Marshallese and Brookhaven personnel to ensure system comparability. 1. Instrumentation The detector chosen for field use by both Brookhaven organizations is a c8 cm diameter, 10 em thick, sodium iodide thallium activated scintillation