Mantle-plume activity has been proposed to explain Neogene and mid-Cretaceous magmatic events, as well as associated tectonism, in West Antarctica; but the arrival time and dimensions of plume influence have been hard to define and are still a subject of debate. Two low-relief erosion surfaces, one in West Antarctica and the other in New Zealand (herein named the Waipounamu erosion surface), provide a way of assessing plume activity by measuring vertical displacements associated with these events. Both surfaces bevel mid-Cretaceous rocks, and both represent prolonged intervals of erosional leveling in a stable tectonic environment. Overlying strata in New Zealand indicate that leveling was near completion in coastal regions by ca. 75 Ma and therefore must have begun around 85 Ma, when New Zealand was beginning to break away from West Antarctica. Fluvial erosion followed by subsidence and marine planation are clearly recorded by these strata, and a similar history seems likely for West Antarctica, accounting for isostatically corrected ice-free bedrock elevations that are well below sea level over much of the region. The absence of uplift at the time of breakup seems incompatible with a plume mechanism for continental breakup. By contrast, the present elevation of the West Antarctic erosion surface records an estimated maximum of ≈ 3 km of tectonic uplift, associated with alkalic volcanism, beginning at ca. 28–30 Ma. We suggest that this event marks the inception of plume activity in West Antarctica. The resulting structure, the Marie Byrd Land dome, defines an area of plume influence that is smaller than the area defined by geochemistry, but is similar in scale to the Yellowstone plume.