The agriculture in Russia (Central and North Region) are characterized by the huge size of the farms, where the average arable field is about ten thousand acres and the average live-stock is about tree-four hundred head per one farm. The transition from the plan to the market economy sharply increase the requirement to quality of the management decision. The Russian agriculture economists call (compel) attention to the modern methods of the farm management. This paper presents the basic directions of the laboratory of economic forecasting at Mari State University (Central Region of Russia). Fuzzy logic, simulating and stochastic models are used in the adaptive methods in agriculture management.