As the largest distribution area of black soil in China, Heilongjiang Province is undertaking important responsibility of food safety for country. However, due to lack of protection, it has become one of those regions undergoing the most serious soil erosion and is gradually losing its basic role of important agricultural production. Soil erodibility index (K-value) is not only an important parameter for evaluating soil's sensitivity to erosion and forcasting soil erosion process but also a key foundation for monitoring and planning of soil and water conservation. In this paper, in terms of soil census data, the characteristics of K-value were analyzed by traditional statistical and geostatistical methods. Results showed that: 1) The soil texture presented significant difference among main soil types; the contents of coarse silt, fine powder and clay obeyed normal distribution, and their spatial correlations were weak. 2) The average K-value of aeolian sandy soils was the largest (0.0281), that of albic soils was the smallest(0.0234), and all K-value of main soil types had weak variability. 3)The average K-value of sandy loam was the largest (0.0281) while moderate clay had the smallest K-value(0.0201). Generally, the average K-value reduced with the increasing physical clay content as well as the decreasing physical sand content, and all soil textures had weak variability but for heavy clay having moderate variability The difference of soil texture either in same soil type or among different soil types was effected by their components content, and this difference was also reflected in K-value. 4) There was a significant linear relationship (R²=0.83) between the average K-value and soil surface thickness and furthermore, the K-value would gently descend when the thickness increased. In other words, the thinner top soil resulted from soil and water loss can cause the increasing K-value and amplify the probability of soil erosion. Thus, the top soil would become thinner and thinner and even disappear. 5) The K-value's distribution characteristics of the areas with various soil erosion types showed obvious difference, however, those regions belonging to same soil erosion type had similar distributions although they were in different soil erosion intensity levels. In water erosion region, the distribution area of different K-values showed an increasing trend in the section from 0.0220 to 0.0241, and got the largest area in the interval from 0.0241 to 0.0245 while a decreasing trend occurred in the zone from 0.0245 to 0.0276. Generally, the distribution area of K-value increased with the increasing of K-value in wind erosion region. The K-value of freezing-thawing erosion region converged in the zone from 0.0258 to 0.0268, and the K-value of engineering erosion region was centered in the zone from 0.0229 to 0.0245. 6) The K-value had a moderate spatial correlation, and showed a gentle linear downward trend from the west to the east, as a distribution of concave parabolic shape occurred in north-south direction. In addition, the maximum K-value mostly appeared in the area of aeolian sandy soil, and the two minimum K-value areas were almost related to the areas of black soils and albic soils, respectively. We also found the distribution features of K-value coincided with the soil and water conservation regionalization well. The K-value grid database produced in this paper would provide basic parameters for soil erosion monitoring and prediction, and controlling of soil and water loss, especially provide scientific basis for the protection and restoration of black soil resources, and impose a positive effect on sustainable grain production safety in the black land. It was necessary to note that the data used in this work were a little old and their soil profile samples were not collected randomly, and thus this shortage of the data may make our results less reliable. Besides, since we lacked newer data, we didn't further analyze relevant spatial and temporal dynamic changes of K-value. [ABSTRACT FROM AUTHOR]