Twenty HASL TLD dosimeter units, each consisting of a carefully selected individually calibrated low background CaFz2:Mn dosimeter and a LiF extruded rod in a shielded container were placed at several different sites on Bikini and Eneu Islands, including many of the spectrometer sites and a few of the NRDL TLD sites. Another 30 of these units were used as controls. Some of these controls were left in Honolulu and Kwajalein on the trip to the atoll and the rest were kept aboard ship. Four were worn by two of us at all times when ashore. The large number of controls allowed us to make very accurate measurements of the average background exposure of the detectors for various stages of the trip. A correction was also made for the background exposure rate to the controls kept aboard ship during the survey. A careful experiment in the laboratory enabled us to account for the self-activation of the CaF2:Mn dosimeters due to *° K in the glass envelopes of the units. The resulting control dosimeter exposures were consistent with the expected expcsure for the time the dosimeters spent on each stage of the trip and their location during that time. The net exposure rates determined from these data for the various field sites are also given in Table 1. Overall, the agreement is quite good considering the small number of measurements, and the results substantiate the overall ccnsistency of the various measurements. It must be kept in mind in interpreting all the TLD results that the results of a single dosimeter placed at a particular location may not be representative of the average exposure rate over a larger area surrounding the dosimeter. The dosimeter may not "see" the same radiation field as an ionization chamber or scintillation detector placed a few feet away. This is especially true of many areas on Bikini Atoll where local hot spots are quite prevalent. Also we are comparing a single measurement at a particular time.with an average exposure per hour determined from a 7 - 10 day exposure. in general, all the TLD results tend to substantiate the ranges and general trends predicted by the data from the active instrumentation.