water sample is the same for each MnO2 cartridge. We found this to be true for *42Pu tracer Pu (III), Pu (IV) and Pu (VI) tested in the laboratory. Collection efficiency is reproducible if the MnO2g cartridges are prepared just prior to use so that any physical damage §o the MnOo surface is minimized. Temperature ranges between 10-30 C. and pH variations between 7-9 seem to have no significant effect on the effi- ciency. The MICE procedure has been used to preconcentrate plutonium from water contaminated by different sources. Fallout plutonium has been separated from surface Pacific Ocean water south of the Farallon Islands, west of San Francisco (37° N 123° W) and near the Marshall Islands (10 N 165 E). Plutonium remobilized to solution from the solid phases >7f the environment has been concentrated from Enewetak Atoll lagoon, reef, and groundwater. To verify the accuracy of the calculated collection efficiency using Equation 1, representative aliquots of 50-200 liters of filtered water from each station were collected and analyzed separately for plutonium by standard radiochemical coprecipitation methods. RESULTS AND DISCUSSION The four types of environmental water samples collected for this study are listed in Tables 1 and 2 along with the sample volumes processed with the MICE technique. Also tabulated for each sample are the 239°240pu concentrations associated with the material removed by the 1 um prefilter (PF), the activity adsorbed on each MnO2 cartridge (A,B), the plutonium concentration in the filtrate determined from Equation 2, the aliquot concentration determined by the radiochemical coprecipitation method (RC), and the collection efficiency of the MnO cartridges. The first point of interest in Table 1 is the good agreement of the 239°240py concentration in the filtrate determined by the MICE and radiochemical method for both the ocean and lagoon samples. Only one (sample E-11) out of twelve samples from these two areas shows a devia- tion greater than the counting error. The plutonium concentration of the filtrate determined by the MICE was calculated by Equation 2 using the measured collection efficiency from each sample. The Pu collection efficiency computed from Equation 1 for the lagoon and open ocean waters (Table 1) varied from 48 to 95% and average 72 + 13%. These differences may be related to sample flow rates, matrices or cartridge loading characteristics, but at this time it is unclear which mechanism is responsible. This variation is noted even in samples collected sequentially from the same area as shown by sample E-l. The result clearly shows that a mean computed collection efficiency would not be adequate for correcting the loss of plutonium in the preconcentration step; 586