The subsidence of young seafloor is generally considered to be a passive phenomenon related to the conductive cooling of the lithosphere after its creation at mid-oceanic ridges. Recent alternative theories suggest that the mantle dynamics plays an important role in the structure and depth of the oceanic lithosphere. However, the link between mantle dynamics and seafloor subsidence has still to be quantitatively assessed. Here we provide a statistical study of the subsidence parameters (subsidence rate and ridge depth) for all the oceans. These parameters are retrieved through two independent methods, the positive outliers method, a classical method used in signal processing, and through the MiFil method. From the subsidence rate, we compute the effective thermal conductivity, k , which ranges between 1 and 7 W m K . We also model the mantle flow pattern from the S40RTS tomography model. The density anomalies derived from S40RTS are used to compute the instantaneous flow in a global 3D spherical geometry. We show that departures from the k =3 Wm K standard value are systematically related to mantle processes and not to lithospheric structure. Regions characterized by k >3 Wm K are associated with mantle uplifts (mantle plumes or other local anomalies). Regions characterized by k <3 Wm K are related to large-scale mantle downwellings such as the Australia-Antarctic Discordance (AAD) or the return flow from the South Pacific Superswell to the East Pacific Rise. This demonstrates that mantle dynamics plays a major role in the shaping of the oceanic seafloor. In particular, the parameters generally considered to quantify the lithosphere structure, such as the thermal conductivity, are not only representative of this structure but also incorporate signals from the mantle convection occurring beneath the lithosphere. The dynamic topography computed from the S40RTS tomography model reproduces the subsidence pattern observed in the bathymetry. Overall we find a good correlation between the subsidence parameters derived from the bathymetry and the dynamic topography. This demonstrates that these parameters are strongly dependent on mantle dynamics.