<p>For the successful functioning of multi-level systems including people, the necessary conditions are coordinated decision-making and taking into account the feedback of systems. The aim of the study is to formalize the decision-making process in multi-level active systems capable of transforming the problem-solving area in a fuzzy information environment. The article proposes a new decision-making algorithm that uses fuzzy models and allows combining the optimal solution to the problem of predicting the behavior of the system and the decision-making process, taking into account the feedback of the systems. The proposed approach allows minimizing the negative consequences of the decision by taking into account the feedback reaction of the systems. The advantages of the approach are demonstrated using the example of treatment patients in clinics. The Bellman optimality principle is reformulated to account for the feedback reaction of systems and the consequences of decision-making. Restrictions on the set of consequences necessary to obtain the acceptable solution are formulated. Decision-making strategies are determined depending on the ratio of the resource capabilities of the system and the time of onset of consequences. The proposed approach is useful, and in some cases, the only acceptable one when solving problems in multi-level active systems. The proposed approach allows improving the quality of decisions. The approach can be applied to a wide class of systems, in which the quality of decision and its implementation depend on the feedback reaction of the systems. Continuation of research is related to studying the problem of sustainability and decision-making in conflict-active systems.</p>