The Pleistocene paleooceanographic and paleoclimate development in the eastern Arctic Ocean is still widely unexplored: this is mainly due to contradictory stratigraphic models and relatively low contents of micropaleontological proxies such as foraminifers, coccoliths and diatoms that are tradionally used to unravel the history of surface and deep waters as well as sea-ice coverage. We have selected ODP Hole 910 for a paleoenvironmental study because presence of biogenic carbonate permits establishment of a relatively continuous stable oxygen isotope stratigraphy on planktic foraminifer Neogloboquadrina pachyderma sin. that is the basic stratigraphic tool in the Pleistocene. Furthermore, the Yermak Plateau is located in the path of inflowing relatively warm North Atlantic waters, and its sediments record the dynamic coupling between the northernmost branch of the Gulf Stream and the Arctic Ocean. Various micropaleontological, sedimentological and organic-geochemical methods are applied on the same sample set to reveal the synchroneity/ asynchroneity of change in the surface water regime and changes in the sediment supply and source areas.The revised chronostratigraphy indicates that the uppermost 20 m of Hole 910A range from Marine Isotope Stages (MIS) 16 to MIS 2. Several stratigraphic age fixpoints support the interpretation of the stable oxygen isotope record that is punctuated by numerous short-term meltwater events. Organic-walled dinoflagellate cysts that have a preservation potential superior to most calcareous and biosiliceous microfossil groups are used to establish a biostratigraphy calibrated vs. the new chronostratigraphy, and to reconstruct the history of surface waters and sea-ice coverage and its interaction with the discharge of freshwater from Eurasia and the growth and decay of the northern Barents Sea ice sheets. Most dinoflagellate cyst samples analysed so far are productive suggesting that Atlantic water inflow into the Arctic Ocean was suppressed only for relatively short periods in the past 600,000 years. Presence of freshwater algae in a number of samples suggests some freshwater supply from the Laptev and Kara seas during certain intervals. Interpretations of the almost monospecific planktic foraminifer assemblages are biased because selective dissolution may have overprinted the paleorecord.The coarse sand fraction (250 to 2000µm) and the number of gravel particles (>2mm/10g) that characterise supply by icebergs do not show much variability, except for a pronounced maximum in MIS marine isotope stage 2 that is probably related to the Last Glacial Maximum (LGM). Minor fluctuations of the gravel content might be related to glacial/interglacial transitions and the onset of glaciations in the adjacent northern Svalbard/Barents Sea.Mineralogical and geochemical parameters show some variability in the Pleistocene siliciclastic sediments of Hole 910A. Carbonate contents are principally low and range usually between 0 and 4%. Maximum values of 8 to 10% were measured in samples from MIS 2. The terrigenous minerals dolomite and siderite clearly dominate in most samples showing that only a rather small portion of bulk carbonate is of biological origin. Small fluctuations are obviously not related to glacial/interglacial cycles. Rock Eval pyrolysis and isotopic signatures of the organic matter (?13Corg) revealed that total organic carbon contents are predominately of terrestrial origin, and are obviously inversely correlated with carbonate contents.Finally, our study will provide reference stratigraphic data sets for interpreting the micropaleontological, sedimentological and organic-geochemical records of the new boreholes that will be drilled on the Lomonosov Ridge (Central Arctic Ocean) in the frame of the Arctic Coring Expedition (ACEX, IODP) in summer 2004.