Abstract-It has been proven recently that using Mobile Agent (MA) in wireless sensor networks (WSNs) can drastically help to obtain the flexibility of application-aware deployment. Normally, in any MA based sensor network, it is an important research issue to find out an optimal itinerary for the MA in order to achieve efficient and effective data collection from multiple sensory data source nodes. In this paper, we firstly investigate a number of conventional single MA itinerary planning based schemes, and then indicate some shortcomings of these schemes, since only one MA is used by them. Having these investigations and analysis, a novel genetic algorithm based multiple MAs itinerary planning (GA-MIP) scheme is proposed to address the shortcomings of large latency and global unbalancing of using single MA, and its effectiveness is proved by conducting the extensive experiments in professional environment.