Hot spots have been widely used as a fixed reference frame for plate motions; however, it should be expected that the underlying plumes are advected in a dynamic mantle. Here a simple numerical model of hot spot motion due to large-scale mantle flow is applied to the Easter, Hawaii and Louisville hot spots. Computations are performed for a range of different buoyant plume rising speeds, ages and locations, mantle density and viscosity structures, etc. It is shown that a westward motion of the Easter hot spot at several cm per year relative to Hawaii and Louisville hot spots is a robust result for a large range of model parameters. In order to obtain an appropriate boundary condition for mantle flow near the Easter hot spot and to find out on which plate to expect the Easter hot spot track for a given time, published isochrons were rotated to the past ridge locations in a reference frame that takes the computed motion of Hawaii and Louisville hot spots into account. Calculations that include flow in the mantle and therefore relative movement of hot spots yield a better explanation of the observed age distribution along the Sala y Gomez ridge than assuming hot spot fixity; at the same time they predict a hot spot track on the Nazca plate that is roughly similar to the shape of the Nazca and Sala y Gomez ridges. Furthermore, the Easter hot spot was in an intraplate location on the Nazca plate during the past 43 Ma, somewhat closer to the ridge between ∼43 and 26 Ma than between ∼26 Ma and now. In contrast, if hot spots are assumed fixed, a location close to the ridge and rather on the Pacific plate between ∼43 and 26 Ma is predicted. Based on morphology and gravity signature it is suggested that the Nazca ridge and the eastern Tuamotu Island Plateau were created by plume material that erupted at the East Pacific Rise while the Easter plume was located sufficiently close to enable such an interaction, whereas the Sala y Gomez ridge was formed above the Easter plume in an intraplate setting. This implies that, if hot spots have indeed moved as computed here, during formation of the Nazca ridge (between ∼43 and 26 Ma) plume material erupted at the ridge several 100 km away from the plume, whereas no volcanism occurred directly above the plume. It also implies that any seamounts on the Pacific plate east of the Tuamotu Plateau are unrelated to the Easter plume.