In this paper, four methods were put forward to predict the transmissibility of an air suspension seat with a seated subject. For characterizing the dynamics of the suspension seat, two of the methods were based on developing a model of the air suspension seat and calibrating the parameters using the transmissibilities of the suspension and complete seat respectively with an inert mass from the experiment. The other two methods substituted the detailed modelling of the seat by two dynamic stiffness connecting in series calculated from two same transmissibilities measured in the experiment. For characterizing the biodynamics of the human body, two of the methods took advantage of the normalized apparent mass from published papers to take place of the human model, while the other two methods made use of one measured seat transmissibility with a subject to deduce the apparent masses of all the subjects. There illustrated good agreement between the experiment and model prediction for all the four methods. In addition, it also exhibited that a large discrepancy can be resulted in, especially in the high-frequency range, if the seat model was substituted by one integrated dynamic stiffness.