neutrons with a fission spectrum,and for Pu-239, similarly fissioned, Ky = 2692. From these and other values determined in the same manner, we conclude that 2900 is a good estimate of Kg for most applications. The detonation products are not, of course, deposited uniformly. The ratio of exposure rate to deposition density has been observed to vary from point to point within the fallout field ,> tending to increase with increasing distance downwind from ground zero, This observation is consistent with the concensus that radiochemical fractionation causes this ratio to decrease with increasing particle size. This problem has been customarily circumvented by using what amounts to an average of this ratio over the region of "local" fallout, where "local" was defined at the convenience of the author. This local averaged K-factor we call Kj. Since local fallout (however defined) represents deposition of only a fraction of the total radioactivity produced by the detonation that produced the fallout, the tatio Ky/Kg has been referred to as the fraction of the activity deposited in the local fallout, or simply "fraction down."' DCPA wants K;, as well as the ratio. However, Iwo additional factors degrade the apparent value of the K-factor. Shielding by small-scale irregularities of terrain leads to a reduction in K,; of about 25% and measuring instruments used in the past have had built-in self-shielding factors that led to another reduction of about 25%. So-called measured values of the K-factor in the literature are nearly always this doubly degraded K-factor, here called Kp. The numerical value of Kj or Kp depends on the definition of local fallout. Three definitions have been used: (1) all deposition out to the distance traveled by particles of a given size, say 45u,, which fall from the top of the nuclear cloud, (2) fallout deposited up to a given time, say H + 24 hours, and (3) the region within a given fallout contour, say 0.5 R/hr at H+ 1 hour. None of these leads to a K-factor completely independent of yield and meteorology, although the first comes closest. We focus here on the third which appears to be the most significant in fallout prediction systems used by DCPA. Empirical determinations of the K-factor make use of the intensity ‘area integral; thus K 1 “WT Ai f do IdA where A is the area (mi?) within the contour of intensity I (R/hr extrapolated from measurements back to H + 1 hour), We is the yield due to fission (kt), and A, is the area within the largest and least intense contour used, E