Demand-responsive Transport (DRT) is a popular topic in the field of transportation. One way to improve the efficiency of the DRT system is to enable pre-booking. The rolling horizon approach can be used to achieve an efficient operation of the DRT system under this context. In this study, a rolling horizon optimizer is implemented in an agent-based transport simulation framework. A sequence of systematic experiments have been performed to study the characteristics of the rolling horizon approach. The results of the experiments provide a general guideline on how to set up the rolling horizon optimizer for a DRT system. With a suitable setup, the rolling horizon approach demonstrates a clear advantage over the existing approaches in the simulation framework.