Non-invasive geophysical techniques offer an interesting alternative to traditional soil sampling methods, especially for estimating spatial variations of soil parameters in the landscape. The spectral analysis of seismic surface-waves (MASW) can be used to determine the vertical shear-wave velocity (Vs) model (i.e., vertical variations in Vs with depth). In our study, MASW soundings were determined at each point in a grid spread over a wind-eroded field plot of 15,600 m2. Vs was then mapped in terms of the thickness and stiffness of the superficial loamy material horizon, which are called ThickLM and StiffLM, respectively. To relate the Vs values to the soil stiffness, cone resistance (Qd) soundings were also performed using a Dynamic PANDA penetrometer. Concurrently, boreholes were used to sample the same horizon for bulk soil density (ρb) measurements. Based on these measurements, large variations in ThickLM were observed. The distribution of Vs values along a 130 m transect allowed for the distinction between two layers corresponding to different mechanical properties. The Vs value of 240 m/s was then used as a limit between the loamy material and the underlying clays. This limit was validated using drilling observations performed on the same transect. Therefore, it was possible to map the ThickLM, which varied between 0.2 and 6.5 m over the entire field. The comparison between the averaged values of Vs and Qd in the loamy material layer showed a significant correlation (R2 = 0.4) such that the mapping of StiffLM was realised from the Vs map and the Vs–Qd relationship. Density comparison between the ρb measured on drill samples and the ρb calculated from Vs was also performed using previously published relationships; however, significant correlations were not observed. The obtained maps of ThickLM and StiffLM were consistent with the expected effects of erosion at the catchment scale and provide indications of historical erosion events. This methodology, which provides a structural and mechanical characterisation of subsurface materials, should help to focus conservation measures to the most threatened areas (i.e., the identification of areas that show a reduced ThickLM and increased StiffLM, which are associated with high soil erosion vulnerability and/or high compaction state).