This study addresses a multi-item, two-stage, lot-sizing problem. The large lot size of transportation from manufacturer to the distribution center (DC) is considered to achieve economy of scale of transportation while for smaller lot size from DC to retailer is utilized for efficiency and responsiveness purpose. Cost minimization, in this study, is a trade-off between transportation cost and inventory holding cost. The developed mathematical model and heuristic algorithm consists of 2 layers which are 1) distribution from a manufacturer to a DC and 2) distribution from a DC to a group of retailers. Furthermore, outsourcing transportation has been considered in order to satisfy some parts of customer demand that do not achieve economy of scale of transportation. Comparison between results of the mathematical model and developed heuristic algorithm demonstrates that the heuristic algorithm results in 2.45% of average deviation from the theoretical bound on the objective and much less computation time than the mathematical model.