Previous studies of Earth rotation perturbations due to ice-age loading have predicted a slow secular drift of the rotation axis relative to the surface geography (i.e. true polar wander, TPW) of order of several degrees over the Plio-Pleistocene. It has been argued that this drift and the change in the geographic distribution of solar insolation that it impliesmay have been responsible for important transitions in ice-age climate, including the termination of ice-age cycles.We use a revised rotational stability theory that incorporates a more accurate treatment of the Earth’s background ellipticity to reconsider this issue, and demonstrate that the net displacement of the pole predicted in earlier studies disappears. This more muted polar motion is due to two factors: first, the revised theory no longer predicts the permanent shift in the rotation axis, or the so-called ‘unidirectional TPW’, that appears in the traditional stability theory; and, second,
the increased background ellipticity incorporated in the revised predictions acts to reduce the normal mode amplitudes governing the motion of the pole. We conclude that ice-age induced TPW was not responsible for the termination of the ice age. This does not preclude
the possibility that TPW induced by mantle convective flow may have played a role in major Plio-Pleistocene climate transitions, including the onset of Northern Hemisphere glaciation.