Considering the extreme variability of plutonium concentrations in vegetation and soil samples taken from Area 13 by other members of the NAEG and the inherent uncertainties involved in estimating vegetation and-soil ingestion rates, the writers of this article felt it would be worthwhile to reexamine the available vegetation and soil data and the theoretical basis for estimating plutonium ingestion rates for herbivores. This led to the development of a quasistochastic model for simulating the ingestion of plutonium by a hypothetical cow or a hypothetical herd of cattle grazing the inner compound of Area 13, unhampered by the realworld restrictions imposed by limited space and vegetation. By means of a simulation model, the area in question can be hypothetically stocked with as many cows as might be desired and for as long a period of time as might be deemed appropriate. Based on well-defined assumptions and making full use of available data, a simulation model based on strictly defined random processes should provide a fair estimate of the mean plutonium ingestion rate as well as an idea of the range of mean values and error terms that might be expected from hypothetical experiments conducted with or without the restrictions imposed on real-world experiments. THE MODEL In previous studies (Martin and Bloom, 1976 and 1977), it was assumed that cattle grazing a plutonium-contaminated desert range would ingest both vegetation and soil and that the total plutonium ingestion rate could be expressed by Tou = 1c, + 1c. (1) where, Ip is the plutonium ingestion rate, pCi/day, u, . : : . T. is the vegetation ingestion rate, g/day (dry weight), and Cy is the plutonium concentration, pCi/g, in vegetation, I. is the soil ingestion rate, g/day, Cc. is the plutonium concentration, pCi/g, in soil. Cattle are assumed to graze randomly within the fenced area (95.5 ha) of the inner compound. What this assumption means, in effect, is that if the area of the inner compound were divided into a grid of very small squares all the same size, each subdivision would have an equal prob- ability of contributing to the intake of vegetation and soil. Individual animals may exhibit a preference for one subarea compared to another or for one plant species compared to another. A whole herd of animals may exhibit seasonal preferences with respect to subdivisions of the fenced area or with respect to plant species, and similar unspecified but nonrandom behavior patterns may apply to the ingestion of soil. The point of the "random-grazing-assumption" is that during the course of a year, the total vegetation and soil consumption by a herd of grazing 485