The change rates of sizes of groups with different educational levels are determined by demographic processes (death rates, ages, migration) and by recruitments from learning institutions and losses caused by moves into higher educational groups. The interrelationship of reproduction of individuals with different training is determined by shifts of population groups from one educational level to another. Total shifts of demographic generations through the educational sphere concurrently with movements from age to age, death rates, and migration is a process of reproduction which determines the structural dynamics of a population's educational potential. This potential's forecasting can be developed only by considering the demographic mechanism of the population's education reproduction in conjunction with its demographic projection. However, first it is necessary to: examine the static structure of population by education, in a form in which it presents itself at each time period; study the temporal patterns of its development; and represent these concepts in a model synthesizing both static and dynamic aspects of population education reproduction. A simulation model of this process has been devised at the Institute of Economics and Industrial Engineering, Siberian Branch of the U.S.S.R. Academy of Sciences, with the computer implementation of its algorithm. This paper summarizes the experience in using this simulation system for the development of long-run predictions of rural population distribution according to education. (NQ)