Conducting effective targeted monitoring of land pollution by oil production and refining is a task that is relevant for all countries, especially for Russia. Much in the efficient solution of this problem depends on the types of pollutant and soil, protection technology. It is important to know the assessments of pollution risks, to localize them. The purpose of the work is system analysis and modeling of land pollution by oil under limited and uncertain data. The hypothesis of analysis and modeling under consideration: "an ecosystem is open, continuously developing and interacting with the environment': Used methods of analysis-synthesis, decision-making, optimization and simulation of systems, an evaluation technique and procedure, as well as expert and heuristic approaches. Multifactorial and uncertainty of the solved problem, which often complicates research and prediction, are proposed. The main result of our work is the methodology of modeling (forecasting) of the state of the land taking into account bifurcations, making managerial (for example, recreational) decisions. The study was conducted with a focus on predicting self-healing of the environment. The results can be used in the development of intelligent systems for assessing land pollution.