Abstract: A global supply chain composed of suppliers, third party logistics service suppliers, and buyers with fuzzy random demands is addressed in this paper. Aiming at total profit maximization for the supply chain, with tariffs and exchange rate considered, a fuzzy random optimization model for order assignment problem is proposed. The model is solved by transforming it into a single objective optimization model for maximizing satisfaction degree. The profit for individual parties is allocated by using Shapely method. A numerical example is used to demonstrate the effectiveness of the model by comparison. It shows that, by maximizing the total profit of supply chain, the total profit can increase by at least 3.47% than that by maximizing the profit of one part or two parts.