in the stainless steel. The instrument response is energy dependent, and data fron the manufacturer indicate an error of as much as 6 to 10% could result if energy corrections are not made to the gross readings®. The RSS-llls used in this study were calibrated at the factory against radium sources whose calibration is traceable to the National Bureau of Standards. The calibration of the instruments was also checked at the Environmental Mon:toring Laboratory (formerly Health and Safety Laboratory) before and after fie.d use. In the report on external exposure for all other atolls surveyed by BNL3, energy dependence corrections were calculated For data from Rongelap and Ronzerik Atolls. The factors needed to compensate the RSS-111 response for energy dependence ranged from 1.01 to 1.05. The mean correction was approxi- mately 1.02. B) Thermoluminescent Survey Lithium fluoride (LiF) thermoluminescent dosimeter chips 1/4-inch square were used, for several reasons. LiF is approximately a tissue equivalent material, and its response is essentially energy independent for photon energies greater than 20 keV up to several MeV. The system is precise to +2% and has a lonz term retention of 5% loss at room temperature for one year. These qualities made the LiF ideal for use in the Marshall Islands. All TLDs were cleaned with analytical grade methanol before departure for the Marshall Islands and prior to analysis. Prior to irradiation, the TLDs were annealed at 400°C for one hour and them at 100°C for 2 hr. After field exposure and before reading, the TLDs were annealed at 100°C for 10 min. In addition to the TLDs exposed in the field at Bikini and Eneu, several sets of TLDs were assembled for use in correcting field measurements for background, fading and air transportation contributions. Several TLDs were annealed and then immediately stored in a lead pig in the BNL analytical counting area. An equal number of TLDs were irradiated co 100 oR and stored with the background TLDs to determine fading losses. Four other TLDs were sent to Kwajalein and stored there in a lead pig to determine in-transit contributions to the response. All TLD results have been corrected for these parameters. The TLDs were calibrated at BNL with 137cs gamma and 90sr/90y betas. Results are directly related to the external exposure and beta absorbed dose that would be received by individuals living on Bikini and Eneu Islands. Because the total response must be differentiated into beta and gamma components, a TLD holder was developed that would eliminate nearly 100% of the 90Y beta of 2.27 MeV (Figure 1). Four TLDs are used per holder. Two are covered by 1100 mg/cm? of aluminum and Mylar which is of sufficient mass density thickness to eliminate beta response; these provide the gamma response. The two other TLDs are shielded by “ 15 mg/cm? Mylar to respond to the total gamma-beta contribution at one meter above the earth's surface. The difference between the