INTRODUCTION The purpose of this paper is to report the results of two field sampling studies into the nature of the variability in Am soil concentrations at NTS. The first study examines the relationship of aliquot size to observed Am soil concentration. The second investigates the pattern of Am variability as a function of the distance between samples in the field. The variability with aliquot size study is important in part because the basic data in field studies are concentrations. Knowledge of the effect of aliquot size on observed concentration and variability is needed to assess the appropriateness of ascribing the observed aliquot concentration to the entire sample. One of the recommendations made at the EPA Workshop (1976) on Soil Collection and Analytical Techniques held in April 1974 was to conduct aliquot studies such as the one discussed here. The variability with distance study was designed to provide information on the interrelationship of Am concentrations over space at the Project 57 (Area 13) site on NTS. This type of information is needed for a statis- tical method called kriging to estimate the spatial distribution of radionuclide concentrations. Additional motivation for these two vari- ability studies is given below. Throughout this paper, 24lam are referred to as Pu and Am, respectively. 2392240py and VARIABILITY WITH ALIQUOT SIZE STUDY MOTIVATION Soil samples collected in the field are typically too large to permit the entire sample to be analyzed for transuranics by wet chemistry techniques. Standard NAEG procedures involve drying the entire sample followed by ball-milling for several hours. Typically one or more 10-g aliquots are withdrawn from the ball-milled sample and analyzed for Pu and Am. The resulting average aliquot concentration (expressed on a per gram basis) is then used as an estimate of the true concentration of the field sample that would be obtained if the entire field sample could be analyzed. Am and Pu in environmental samples are often known or assumed to be in particle form, in which case variability between aliquots is inevitable due to random differences in the number and size of particles in different aliquots. This randomness will exist even though ball-milling has been conducted. Incomplete ball-milling or mixing may, of course, contribute additional variability. Presumably, a complete dissolution of the entire sample into liquid form would result in truly homogeneous aliquots. 407