Abstract: By using Fiber Bragg Gratings (FBG) sensors as strain sensors and developing an effective algorithm of the real-time and in-situ loading recognition, an online load monitoring (OLM) system has been developed. The system was integrated into the software platform of a dynamic FBG wavelength demodulator (sm130-700, MOI) based on a graphical programming language, Labview 8.0, and then applied to a pterygoid cantilever to carry out the load recognition experiment. The experimental results indicate that the present OLM system is capable of identifying both the location and magnitude of load with a high precision and has the satisfactory fault-tolerance.