The Prydz Bay represents the third largest shelf area in Antarctica and is bounded to the south by the outlet of Lambert Glacier. The latter drains about 20 % of the East Antarctic Ice Sheet. Fluctuations in the extent and configuration of shelf ice and sea ice are known to be of wide influence on bottom-water formation in Antarctica. They are thus of global impact as Antarctic Bottom Water (AABW) is a driving force of Thermohaline Circulation. We apply sedimentological, mineralogical and geochemical methods on sediment cores retrieved from Prydz Bay, MacRobertson Shelf, the continental slope off MacRobertson Land, and from the East Kerguelen Drift deposit to reconstruct past variations in the formation of bottom-water and calving of icebergs.A depth transect of sediment cores from the continental slope off MacRobertson Land enables us to distinguish between periods of contourite and/or turbidity current activity and ice-rafting events. We use this information to draw conclusions on the advance and retreat of ice shelves and to deduce the glacial history in the Prydz Bay hinterland during the Pleistocene. Dating of the sediment records was done with the help of palaeomagnetics and diatom stratigraphy. Though interpretation of the sediment records is complicated by hiatuses and difficulties in age determination we identified two major phases of glacial retreat within the last 1.3 Ma. These episodes coincide with results from a drill hole through Prydz Mouth Fan and with exposure age data from Prince Charles Mountains derived from cosmogenic nuclides. The first and obviously very strong glacial retreat took place between ~1.0 - 0.76 Ma, beginning in the time of Jaramillo palaeomagnetic reversal and pronounced interglacial MIS 31. A later retreat started about 120 ka ago after the pronounced interglacial MIS 5.5. Possible changes in the mineralogy of ice-rafted material with respect to the earlier glacial regression tentatively hint to shifts in the provenance of IRD. In contrast to findings gained from exposure data from Prince Charles Mountains we didnt find evidence for glacial retreat at ~450 ka.The sediment core from East Kerguelen Drift covers the time span beginning MIS 6. Grain size analysis and statistical interpretation of the data in combination with mineralogical results hint to strongest bottom-water activity during MIS 5.1 which contradicts results from the Antarctic realm published so far. We speculate that during warm interglacials (e.g. MIS 5.5) bottom-water production is reduced due to less formation of sea-ice and associated brine rejection. This assumption is supported by findings from a sediment core recovered from Burton Basin, MacRobertson Shelf. In this sediment basin water convection obviously decreased during the mid-Holocene optimum as a result of either weakening of Antarctic Coastal Current or reduced brine rejection in the area of Cape Darnley Polynia.