Currently, there is an urgent need for environmental studies of agricultural production. This is especially true for cattle farms. In the situation of uncertainty of the initial information, the most convenient research method is mathematical modelling based on the fuzzy-probability approach. An algorithm for extracting and formalizing expert knowledge and experience by analytical expression is presented. In contrast to foreign studies, the joint influence of seven variables in one model is considered. At the same time, the factor space systematically represents the phenomenon under study and contains not only quantitative, but also qualitative variables. The impact of only technological factors was assessed, so economic indicators were not considered. An example of constructing a fuzzy-probability model of the environmental sustainability of a cattle farm in a seven-factor space is given. The validity of the model is confirmed by the factual data for three farms. Quantitative estimates of the degree of influence of both the variables themselves and their interactions have been obtained. It is shown that the influence of the number of head on the farm and the fields is about twice as significant as the influence of the applied technologies. A numerical experiment is carried out and illustrated on the basis of an adequate fuzzyprobability model. At the same time, it was found that with a significant increase in the number of cattle, the ecological stability of the farm decreases on the scale of the attribute from the level of "high" to "medium".