Abstract: In order to effectively manage the schedule and cost of complex construction pro jects, an algorithm is developed to optimize task flows. This algorithms uses De sign Structuring Matrix as the basic analysis tool, and optimizes the task flow through reorganizing the tasks execution sequence. As learning curve will play a n important role in the completion of complex project, this paper also takes thi s factor into consideration. The simulation results in the case study indicates that it is quite effective for projects with high task interactions, and the lea rning curve will pose a significant effect upon project cost and schedule.