Subminiature fixed-wing unmanned aerial vehicles (SUAVs) present an enormous potential for low-altitude exploration applications. In order to develop a robotic SUAV and realize high level of autonomy, the flight motion model of the platform with wingspan of 1.8 m is studied thoroughly. Based on the rectilinear trim flight and small perturbation theory, the linear parametric model is investigated. The flight experiments are designed, and the flight data measurement system is developed to collect the input and output data. Through outliers processing, data smoothing, data fusion, etc, all the required data for parameter identification are figured out. As the estimation method, improvements of GA are detailed, and the comparison of model verification indicates that improved GA has prominent effects in flight motion model parameter identification of SUAV.