Abstract: Wireless structural health monitoring research has drawn great attention in recent years from various research groups. While sensor network approach is a feasible solution for structural health monitoring, the design of wireless sensor networks presents a number of challenges, such as adaptability and the limited communication bandwidth. To address these challenges, we explore the mobile agent approach to enhance the flexibility and reduce raw data transmission in wireless structural health monitoring sensor networks. An integrated wireless sensor network consisting of a mobile agent-based network middleware and distributed high computational power sensor nodes is developed. These embedded computer-based high computational power sensor nodes include Linux operating system, integrate with open source numerical libraries, and connect to multimodality sensors to support both active and passive sensing. The mobile agent middleware is built on a mobile agent system called Mobile-C. The mobile agent middleware allows a sensor network moving computational programs to the data source. With mobile agent middleware, a sensor network is able to adopt newly developed diagnosis algorithms and make adjustment in response to operational or task changes. The presented mobile agent approach has been validated for structural damage diagnosis using a scaled steel bridge.