for this area. Krey and Hardy (1970) reported about plus or minus 25 percent uncertainty. The variability of the upper l-cm technique (average coefficient of variation of 42 percent, range of 18 to 69) is similar to the variability indicated by duplicate sample analysis (estimated C of 29 percent, range of up to 60 percent). Thus, it appears that a significant, if not major, fraction of the variability in duplicate soil samples is related to the variability resulting from aliquoting samples for analysis and This is in agreement with the the actual variability in analysis. conclusions of Krey and Hardy (1970). Many of the lower-level activity samples (Sites I and III, Table 1) had analytical counting errors of 20 to 40 percent (2 sigma). These two sites generally had higher C's than the other sites. The results of this project indicated about 30 percent (range of 30 to 50) of the plutonium-239 in the upper 10 cm of soil was in the upper 1 cm, with about 86 percent in the upper 5 cm. This is a somewhat higher fraction than indicated by Krey and Hardy (1970) for the Rocky Flats area, but the range of values overlaps. It does appear to be at variance with the results of Poet and Martell (1974), which indicated 80 percent of the plutonium-239 in the upper 10 cm was in the top 0.3 cm. The activity per gram values (averaged for each site) for the 0- to l-cm results ranged from about two to five times those for the O0- to 5-cm technique, indicating, as expected, that the 1 em of soil is generally greater than that below ed, the activity per unit area for the 0- to l~cm about 23 to 62 percent (average of 35) of that in activity in the surface 1 cm. Also, as expecttechnique ranged from the O- to 5-cm samples. From this, it is observed that O- to 5-cm samples can be used to conservatively estimate the activity per unit area for the upper 1 cm. But, results from samples taken from a depth of less than 1 cm cannot be directly used to estimate the plutonium-239 per unit area in the surface 1 cm of soil, which is specified by the proposed EPA guidance (EPA, 1977a). It appears there is probably some compromise in the variability of sample results associated with collecting samples related to the surface 1 cm of soil versus the top 5 cm, using the techniques applied in this study. The average C for the l-cm technique was 40 percent greater than that for the 5-cm technique. But, given the greater applicability of plutonium results from 0- to l-cm versus O- to 5-cm samples for resuspension estimates, the increased uncertainty would appear to be accept~ able for programs concerned with assessing potential airborne hazards. The EPA proposed transuranium guidance (EPA, 1977a) specifies the plutonium per unit area in the surface 1 cm. In summary, it is concluded that the surface l-cm technique used in this study can be used to gener- ally characterize an area, 668