An optimal dispatch strategy is crucial for an isolated wind-diesel power system to save diesel fuel and maintain the system stability. The uncertainty associated with the stochastic character of the wind is, though, a challenging problem for this optimization. In this paper, a dynamic multi-stage decision-making model is proposed to determine the diesel power output that minimizes the cost of running and maintaining the wind-diesel power system. Optimized operational decisions for each time period are generated dynamically considering the path-dependent nature of the optimal dispatch policy, given the plausible future realizations of the wind power production. A numerical case study is analyzed and it is demonstrated that the proposed stochastic dynamic optimization model significantly outperforms the traditional deterministic dispatch strategies. (C) 2015 Elsevier Ltd. All rights reserved.