Objectives The aim of the work is the use of various models for optimizing the production of agricultural products for management using empirical data and expert estimates.Method It is proposed to use individual or group averaged and interval values of production and economic parameters as expert estimates for the above linear programming problems under uncertainty. The use of averaged or interval expert evaluations is associated with the features of applied extreme problems and providing models with initial data.Result Models are proposed for optimizing the production of agricultural products and harvesting wild food resources, for the implementation of which it is necessary to use expert estimates that supplement empirical data. The first task of optimizing the crop structure used interval values of the coefficients proposed by experts that characterize the timeliness of implementation technological operations. In the second task related to the planning of procurement of wild-growing products, interval expert estimates reflect fluctuations in the yield of wild plants, labor costs, and cost indicators of production. In the third task, an individual expert assessment of the effect of pollution on annual, perennial grasses and crops was used.Conclusion Optimization tasks for agricultural production, taking into account expert assessments, have been implemented for two agricultural enterprises in the Irkutsk region and the Bratsk region, which has significant reserves of wild food resources.