The Antarctic Ice Sheet is a key feature of the global climate engine today, and has been so for most of its 35 Myr or longer history. It influences global circulation (mainly through bottom water production), eustatic sea-level change, biological production and albedo. And yet the details of that history are poorly known, despite two decades of measurement and interpretation of lowlatitude ice-volume proxies. The most effective of those proxy measurements, oxygen isotopes and sea-level change, are ambiguous, and disagree. Recently, a way out of this impasse has emerged, that is technically difficult but much more direct. It involves sampling and dating sediments transported beneath the grounded ice sheet and deposited seaward of the grounding line around the Antarctic margin. We now appreciate that the ice sheet “drains” mainly by rapid flow in ice streams that slide on a shearing bed of diamict. Over the life of the ice sheet, those glacially-transported sediments have formed progradational wedges on the outer continental shelf. They, and their derived sediments redeposited in drifts on the upper continental rise, should therefore contain a record of ice sheet advance to the continental shelf edge. The prograded wedge is essentially unsorted making it difficult to recover, and the topsets are prone to subsequent erosion. The drifts have formed by more continuous deposition of sorted silty clays that are easier to recover but less direct, needing clues from the wedge to aid interpretation. The two depositional environments are complementary. Additional useful features of the Antarctic margin are the deep basins eroded on the inner continental shelf during glacial maxima, which preserve an expanded Holocene record of climate change.