Large igneous provinces (LIP) are often found in close temporal and spatial proximity with continental breakups, supporting the model, that an arriving mantle plume produces large amounts of melt and has a massive influence on the breakup process. The South Atlantic is a classical example, with flood basalts on both adjacent continents and a paired age progressing ridge system connecting them with the current hotspot location at Tristan da Cunha. To estimate the influence of the plume on the preexisting continental crust, a large-scale geophysical experiment was conducted in 2011 at the intersection of Walvis Ridge with the African continent. We present four P-wave velocity models derived from seismic refraction data. One extends 430 km along the ridge crest and continues onshore to a total length of 730 km, while the other three crossing the ridge perpendicular: one (480 km long) far offshore in the oceanic regime, one (600 km) close to shelf break and the last one (400 km) onshore. Crustal velocities beneath Walvis Ridge range between 5.5 km/s and 7.0 km/s, which are typical velocities for oceanic crust. The crustal thickness, however, is approximately three times larger than of normal oceanic crust: 17 km in the western part increasing to 22 km towards the continent. The continent ocean transition is characterized by 30 km thick crust with a high velocity body (HVB) in the lower crust and seismic velocities up to 7.5 km/s. The western extend of the HVB is to a similar distance from shore as for HVBs observed south of Walvis Ridge. In contrast, the eastern boundary lies well within the continental domain, at the 40 km thick crust of the Kaoko fold belt. Here, the variation of seismic velocities indicates that hot material intruded the continental crust during the initial rifting stage. However, beyond this relatively sharp boundary (40 km wide), the remaining continental crust seems unaffected by intrusions and the root of the Kaoko belt is no eroded. The cross-profile indicates that Walvis Ridge might be broader than its topographic expression and that the northward lying seamounts are part of the ridge. A HVB can only be found at the northern flank of the ridge, but not at its base. We conclude, that the postulated arriving plume head did not modify the continental crust on a large scale, but was a rather regional anomaly.