Vegetation height and vertical structural profile constitute the critical parameters for the  forest biomass and carbon cycle model, and the polarimetric SAR Interferometry (PolInSAR) technique makes quantitative vegetation structural parameter inversion possible. In this paper, a dual-baseline twice-fitting method was used to extract the vegetation height and then, as prior information, a vegetation vertical structural profile was further estimated by expanding the polynomial of the vertical structural function. Finally, the simulated data and real data were used to validate this dual-baseline technique. The experimental results demonstrate the effectiveness and feasibility of the forest vertical structural profile inversion using interferometric coherence variation under different polarization conditions.