Our knowledge of glacial history of the western (Norwegian) part of the Barents Sea has greatly improved during the last decades, notably due to the high-resolution multibeam swath bathymetry data. In contrast, published seafloor data from the eastern part of the Barents Sea and the Kara Sea are much more sparse. This study presents new geophysical/geological evidence for reconstructing glacial dynamics of the eastern part of the Barents-Kara Ice Sheet during the Last Glacial Maximum and subsequent deglaciation. Archival data used in this study include more than 300,000 km of sparker and high-resolution Parasound profiles verified by boreholes drilled with continuous core recovery to 50-100 m below sea bed. This dataset was used to construct continuous geological cross-sections and a series of maps, including detailed bathymetry (in 10-m isobaths) and sediment thickness maps of major seismo-stratigraphic units. Based on the bathymetric and sediment thickness data we map megascale glacial lineations, drumlin-like ridges up to 50 m high and subglacial channels up to 100 m deep, as well as accumulations of glacial deposits (basal, lateral and end moraines) and ice-proximal acoustically transparent bodies (ATBs). Spatial and stratigraphic analysis of these bedforms enables us to put forward a new hypothesis that ice moved on the shelf from the Arctic Ocean along the Saint Anna Trough (SAT). Further south, near the northern tip of the Novaya Zemlya islands, the ice flow split into three major lobes moving to the southwest into the Barents Sea and to the south and southeast into the Kara Sea. Deglaciation in the study area progressed with several ice stillstands and subsequent readvances marked by end-moraines and accumulation of ice-proximal sediments. During deglaciation events, when the SAT became ice free due to iceberg calving, the ice flow reversed its direction toward the SAT, forming a fluting and a massive ATB on the western SAT slope. The exact timing and mechanisms of the ice transgression(s) from the Arctic Ocean are not well understood. Additional high-resolution data such as multibeam bathymetry surveys are needed to verify the spatiotemporal distribution of glaciogenic bedforms, and glaciological modeling is required to comprehend the ice dynamics and put it in the pan-Arctic context.