The statistical characteristics of the contamination fields and the measured dose rates of gamma radiation in the air for the territories contaminated by the Chernobyi accident are of interest from the point of view of assessment of the absorbed dose for the population. The time dependence of the dose rate and the composition and characteristics of the contamination for individual territories of Belarus and Russia were considered earlier [1]. Examined here are the statistical characteristics of the contamination fields and dose rates in the Bryanskaya Oblast of Russia and the adjacent Gomel 'skaya and Mogilevskaya Oblasts of Belarus. Data on soil-sample 137Cs concentrations in the "Store" data base of the "Typhoon" Scient i f ic-Industr ia l Organization as well as data on the dose rate for the autumn of 1990 are analyzed. The 137Cs soil contamination and dose rate were determined simultaneously from samples or in a special manner for example, by going door-to-door in contaminated populated areas. The dose rates were estimated for individual parts of the farms (inside the house, on the street at the farm entrance, in the kitchen garden, the site at which the soil sample was taken). The soil samples were taken in the sodded regions using a standard sample cutter (a ring 140 mm in diameter and 50 mm high; the area of a unit sample was 0.0152 m 2) and in the gardens (plowed areas) by a soil auger to a depth of 200 mm. Mechanism of Formation of Statistical Characteristics of Soil-Contamination Distribution. Contamination can be formed by a small number of aerosol particles or rain drops that have landed on the area of the sample cutter. Modeling by the Monte Carlo method of the fallout of aerosol particles with a size of 0.1-5.5 mm (the spectrum of rain drops in this range has been described by Steiner and Waldvogel [2], for example) shows that the areal distribution of the mass of the particles is welldescribed by a log-normal distribution or a similar exponential distribution that has been truncated on the side of small particle size. The statistical density distributions are estimated by measurement of the activity of soil samples with a small area. Therefore, the distributions must be asymmetrical or sharply limited from the small-value side (Weibull or log-normal distributions). Statistical Data Processing. For statistical processing, the data on contamination density (P, Ci/km 2) and measured dose rate were grouped by various methods. Detailed data for individual populated areas were examined first (the distance between measurement and sampling points was several tens of meters). Then data averaged over the territory of populated points of a selected administrative region were used (the distance between sampling points was several kilometers). The distribution parameters for these groups of data were different, although the type of distribution was the same. For the grouped data, the type of statistical distribution was determined from an a priori set (gamma, Weibull, or log-normal distribution). The quality of the data description was verified by the Z 2 test. In addition, distribution-type-independent (robust) statistics the median and interquantile range were employed for a general description of contaminations different scales on the territory. Quartiles and decries, i.e., the values between which 50% and 80%, respectively, of all values were included, were used as quantiles. Contamination-Density Distributions for Individual Populated Points and Larger Territories. It follows from Fig. 1 that the contamination distributions over territories of populated points with high and relatively low contamination are similar. In all cases, the contamination distribution is nearly log-normal (as use of the g 2 test shows). As can be seen from Table 1, the characteristic of populated-point contamination used at the present time the mean 137Cs contamination differs considerably.