Over the last three decades, the Bering Sea has undergone dramatic changes in its physical and biological environment due to anthropogenic warming, accompanied by a dramatic shift in sea-ice cover, thickness, and duration of the ice season. These recent observations have increased the interest of the paleo-community in reconstructing the past variability of sea-surface characteristics in this region. The main objectives of this study were the millennial-scale reconstruction of the variability of surface-water conditions and terrigenous input in the subarctic Pacific and Bering Sea over the last glacial-deglacial-Holocene time interval (25 ka) in order to investigate their relationship to global climate change. For this purpose, changes in organic geochemical composition in surface sediments and sediment cores, collected during the SO202- INOPEX RV Sonne and SO201-KALMAR R/V Sonne cruises were investigated. The main results of this thesis are presented in three manuscripts. In order to determine the applicability of specific biomarkers, in view of future paleoclimatologic and paleoceanographic reconstructions in the subarctic Pacific and the Bering Sea, the first manuscript was dedicated to the reconstruction of modern sea-surface characteristics, i.e., sea-surface temperature (SST) and sea-ice cover in this area. Based on hydrogen index values and the distribution of long-chain n-alkanes and specific sterols in surface sediment, we show that different organiccarbon sources prevailed in the study area. In the Bering Sea, organic matter has a predominantly marine origin, caused by high primary production, whereas in the North Pacific, organic carbon originates mostly from terrestrial higher plants, probably related to dust input from Asia. The results from the alkenone-based SST reconstruction demonstrate that the Sikes et al. (1997) calibration seems to be more accurate and matches the summer SSTs in the eastern North Pacific and the Bering Sea better than the Müller et al. (1998) calibration. In this study, we also show that the distribution of the novel sea-ice proxy IP25 in surface sediments mirror the modern spring sea-ice distribution and demonstrates the potential of this proxy to track past variations in sea-ice cover in the study area. In the second manuscript we use IP25 abundances, alkenone-based sea-surface temperatures, diatom and biogenic opal data from three sediment cores from the western North Pacific and western Bering Sea to reconstruct the variability of sea-ice extent during the past 18 ka. In general, there is a very good correlation between the biomarkers-based and the diatoms-based sea-ice records. The results demonstrate that a dominantly permanent sea-ice cover prevailed in the western Bering Sea during cold periods (Heinrich Stadial 1 and Younger Dryas), whereas reduced sea-ice or ice-free conditions existed during warmer intervals (Bølling/Allerød and Holocene). Warm intervals of reduced sea ice coincide with increased biogenic opal, indicating increased primary production. In the last manuscript, the millennial-scale of sea-ice reconstruction was extended to include the northeastern Bering Sea and the eastern and western subarctic Pacific. The results show that an extensive sea-ice cover prevailed over large parts of the subarctic Pacific and the Bering Sea during the LGM. The deglaciation-Holocene time interval is characterized by rapid sea-ice advance and retreat. During cold periods (Heinrich Stadial 1 and Younger Dryas) seasonal sea-ice cover generally coincides with low alkenone SSTs and low primary productivity. Conversely, during warmer intervals (Bølling/Allerød and Holocene) predominance of reduced sea ice or ice-free conditions are generally associated with increase in alkenone SSTs and primary productivity. However, in the northern Bering Sea continental shelf ice-free conditions prevailing during the Holocene Thermal Maximum shifted to marginal sea-ice conditions at the onset of the Mid Holocene. In summary, the work in this thesis demonstrates that sea-ice extent in the subarctic Pacific and the Bering Sea was highly variable during the last 25,000 years. The variability can be explained by a combination of local factors (e.g. solar insolation), as well as global climate anomalies (e.g. Bølling/Allerød and Younger Dryas) and sea-level changes controlling the oceanographic circulation between the subarctic Pacific and the Bering Sea.