The paper analyses the influence of the foreseen climate changes on uncertainty of agricultural production in Russia as a whole and at the federal subjects level. The methodology relies on a linear computable mathematical model that uses a non-parametric representation of the production frontier in a form that accounts for uncertainty, which is inherent to agricultural production. Uncertainty is measured by the standard deviation to mean ratio of a given agricultural output. Under the conditions of the global warming the uncertainty of total agricultural output decreases, while it increases for each studied output (grain, milk, meat, remaining outputs). As for the federal subjects, they demonstrate the diverse changes of uncertainty. Their direction depends on multiple factors including the local climate and its expected change, technical level of production, the capacity of a more efficient usage of resources.