This study focuses on the identification of aerodynamic coefficients through airdrop data based on a known six-degree-of-freedom (6-DOF) model. Since the precise modelling of the parafoil system is a prerequisite for design of guidance, navigation and control systems, the accurate calculation of aerodynamic coefficients is vital. This paper first presents a rigid 6-DOF nonlinear model and simplifies it by linearizing the equations around the equilibrium point, which facilitates the acquisition of aerodynamic parameters next. Then, the recursive weighted least square method is applied to identify roll and yaw coefficients from airdrop test data. At last, simulations in the wind environment are implemented to analyse the dynamics of gliding and turning. The airdrop experiment also verifies the effectiveness of the dynamic model and the accuracy of identification.