Abstract: The porosity of human cortical bone will increase when osteoporosis occurs. To study the relationship between cortical bone parameters（thickness, transverse wave velocity, longitudinal wave velocity） and osteoporosis,the finite element method is used to simulate the cortical bone with different porosities（0∶3%∶27%） in this paper.By using a 3-period Gaussian envelope sine wave as excitation, the acquired ultrasound guided wave signals are processed by the two-dimensional Fourier transform and Burg algorithm to get the experimental dispersion trajectories, and then by grid searching to match them with the theoretical dispersion database established by the Floquet-Bloch theory to get the cortical thickness, transverse velocity, and longitudinal velocity. The results show that the cortical bone thickness inversion is accurate, and the cortical bone porosity is negatively correlated with the transverse and longitudinal velocities. The transverse velocity sensitivity is 19.0% and the longitudinal velocity sensitivity is 5.5%, which indicates that the transverse velocity is more sensitive to the porosity and has greater clinical diagnostic potential for osteoporosis. Experimental signals are also collected from six bovine cortical bone plates. The inversion results show that average relative error of the cortical thicknesses is 4%, and the experimental dispersion trajectories match well with the theoretical dispersion curves, therefore the feasibility and accuracy of the proposed inversion algorithm in this paper is verified. The research in this paper has application potential for ultrasonic detection of osteoporosis.