Crucial component of the successful performance of all participants of logistic process as an economic system is maintenance of an optimality of functioning of all industrial and economic criteria of the given system. The purpose of the given work is stochastic modelling of the stability of transport logistics system (TLS). The process of deterministic and stochastic modelling of TLS stability is considered in the conditions of uncertainty using bootstrapping and dynamic programming method by Bellman. It is well known that random character may be an attribute inherent to the majority of economic events and processes of TLS. Modelling of logistical processes in conditions of uncertainty is complicated due to the lack of trustworthy information describing the conditions of uncertainty, and also in view of the random character of occurrences of deviations in the course of processes researched. Ignoring the evaluation of effect of certain factors during the research of processes occurring in transport logistics system frequently expands “zone of risk”, entailing mistakes and discrepancies in the real time situation that, in turn, may finally result in significant material losses.