Error in the estimation of the ground interferometric phase and that in the forest scattering model will bring about errors of the inversion of the vegetation height. Aiming at this problem, a new inversion method based on the BP neural network was proposed. The new method directly fits the nonlinear mapping relationship between the complex polarimetric correction coefficients and the vegetation height, so it can reduce the error caused by the error in the estimation of the ground interferometric phase, and that caused by the scattering model error. The new method has better performance than the three-stage vegetation height inversion method, and the simulated results validate the superiority of this new method.