Load identification plays an important role in structural health monitoring. A method to identify load for a cantilever beam based on dynamic strain measurement by FBG (fiber Bragg grating) sensors is presented to facilitate the control over the system during structural health monitoring. The algorithm is based on Kalman filter, using the strain measured by FBG sensors as observed signal, and the gain matrix, the residual innovation sequences and covariance matrix generated by Kalman filter to estimate the load in real time through least squares algorithm. The proposed load identification method based on FBG sensors is a recursive method, which means that recent measurement value and previous estimated value need to be kept in storage. This will save considerable memory and greatly decreases the system burden. The proposed method is based on Kalman filter, and this can be helpful for system control by using optimal control theory after identifying load. To prove the effectiveness of the proposed method, numerical simulations and experiments of the beam structures are employed and the results show that the method has good stability and real-time capability.