The negative impact of global and regional climate changes upon the crop yields leads to the violation of the crop production stability. The development of reliable methods for assessment of the climatic factors by the reaction of the crops to them in order to minimize the impact of climatic stresses upon the sustainability of food systems is an urgent scientific task. This problem was studied on the example of growing corn. A mathematical analysis of the main meteorological indicators for 16 years of research has been performed on the basis of which the frequency and direction of the occurrence of atypical and extreme weather conditions in various periods of the corn vegetation season were established by the coefficient of significance of deviations of the weather elements from the average long-term norm. It has been proved that the probability of occurrence of such weather conditions in the period from April to September is 38–81% in terms of the average temperature of the month, and 31–69% in terms of precipitation. By using the information base of the corn yields in a stationary field experiment with the gradations of factors: A (the fertilizer option) – A1-A12, B (the crop care method) – B1-B3, C (the hybrid) - C1-C7, the most critical month of the corn ontogeny was established when the weather has a decisive influence upon the formation of the crop. With the help of the correlation-regression analysis it was proved that the corn yield most significantly depends on the average monthly temperature in June, and for the hybrids with FАО 200–299 – on the amount of precipitation in the month of May. The obtained mathematical models make it possible to predict the yield of corn at a high level of reliability depending on the indicators of the main climate-forming factors in June, that is, even before the flowering of the plants (before the stage of ВВСН 61).