Throughout the latter half of the Pleistocene epoch of Earth history, beginning ~900 kyr ago, the climate system has been dominated by an intense oscillation between full glacial and interglacial conditions. During each glacial stage, global sea level fell by ~120 m on average, as extensive ice sheets formed and thickened on the surfaces of the continents at high northern (primarily) and southern latitudes. Within each cycle this glaciation phase lasted ~90 kyr and was followed by a much more rapid deglaciation event which terminated after ~10 kyr and which returned the system to the interglacial state. The period of the canonical glacial cycle has remained very close to 100 kyr since its inception in mid-Pleistocene time. Because of the magnitude of the mass that was redistributed over the surface of the Earth during each such glacial cycle and because of the viscoelastic nature of the rheology of the planetary mantle, these shifts in surface mass load induced variations in the shape of the planet that have been indelibly transcribed into the geological record of sea level variability. Indeed, the geological, geophysical, and even astronomical signatures of this process, which is continuing today, are now being measured with unprecedented precision using the methods of space geodesy and have thereby begun to provide important new scientific insight and understanding, both of the interior of the solid Earth and of the climate system variability with which the ice ages themselves are associated. In this article my purpose is to bring together, in a single review, an assessment of where we currently stand scientifically with regard to understanding both of these aspects of the ice ages. Although the discussion will not address in any detail the fascinating issue of ice age climate, since this topic is sufficiently complex of itself to require a detailed review of its own, I will nevertheless attempt to briefly summarize the current state of understanding of the physical processes that are responsible for the occurrence of the ice age cycle, by way of providing a more complete context in which to appreciate the main lines of argument that will be developed. ¿ 1998 American Geophysical Union