This paper aims to solve an order allocation problem in supply chain management (SCM). In this problem, the price fluctuation over consecutive time periods is known beforehand. Based on that assumption, clients can arrange their orders in advance to minimize the costs incurred in SCM, e.g., purchasing cost or inventory holding host. However, order allocations are subject to demand volumes, time-varying costs, suppliers’ capacities, inventory levels, and inventory costs, etc. It is difficult to make a correct decision immediately. With so many factors affecting the decision-making, traditional methods (e.g., dynamic programming) is very time-consuming. Therefore, a genetic algorithm cooperated with a local search is proposed. The experimental results show that the problem can be solved efficiently and near optimally.