Cortical bone loss is not fully assessed by current X-ray methods, and there is an unmet need in identifying women at risk of osteoporotic fracture who should receive a treatment. The last decade has seen the emergence of ultrasound axial transmission techniques to assess cortical bone. Recent axial transmission techniques exploit the multimode waveguide response of long bones such as the radius. A recent ex vivo study by our group evidenced that a multimode axial transmission approach can yield simultaneous estimates of cortical thickness and stiffness. The aim of the present work is to move one step forward to evaluate the feasibility of measuring multimode guided waves in vivo and to infer from it cortical thickness. Measurements were taken on the forearm of 14 healthy subjects with the goal to test the accuracy of the estimated thickness using the bidirectional axial transmission method implemented on a dedicated 1-MHz linear ultrasound array. This setup allows determining in vivo the dispersion curves of guided waves transmitted in the cortical layer of the radius. An inverse procedure based on the comparison between measured and modeled dispersion curves predicted by a two-dimensional transverse isotropic free plate waveguide model allowed an estimation of cortical thickness, despite the presence of soft tissue. The cortical thickness values were validated by comparison with site-matched estimates derived from X-ray high-resolution peripheral quantitative computed tomography. Results showed a significant correlation between both measurements (r 2 = 0.7, p < 0.05, RM SE = 0.21 mm). This pilot study demonstrates the potential of bidirectional axial transmission for the in vivo assessment of cortical thickness, a bone strength-related factor.