Arctic sea-ice extent has been declining rapidly throughout the past decades. To understand the past temporal and spatial sea-ice variability is of significant importance for predicting the future development. Within this thesis, the recently developed sea-ice proxy IP25, exclusively produced by diatoms in sea ice, has given a new insight into the reconstruction of sea ice across major parts of the Arctic Ocean during modern time and the Last Glacial Maximum (LGM). In the first study, determination of IP25, phytoplankton-derived biomarkers (brassicasterol and dinosterol) and terrigenous biomarkers (campesterol and ß-sitosterol) in surface sediments from the Kara and Laptev seas is used to estimate modern spatial (seasonal) sea-ice variability and organic-matter sources. C25-HBI diene and triene were determined as additional paleoenvironmental proxies in the study area. Furthermore, a combined phytoplankton-IP25 biomarker approach (PIP25 index) is used to reconstruct the modern seaice distribution more quantitatively. Neither IP25 nor PIP25, however, show a clear and good correlation with satellite sea-ice distribution due to the complex environmental conditions in our study area. Differences in the diene/IP25 and triene/IP25 ratios point to different sources of these HBIs and different environmental conditions. The diene/IP25 ratio correlates positively with sea-surface temperature, and thus might be used as a potential SST index. Further studies are, however, needed to validate this index. Furthermore, in the second study, a comprehensive data set of these biomarkers was produced using surface sediment samples from the Central Arctic Ocean proper (>80°N latitude) and the Chukchi Plateau/Basin. In addition, published data from other Arctic and sub-Arctic regions were added to the new data to generate an overview distribution map of IP25 across major parts of the Arctic Ocean. The phytoplankton biomarkers brassicasterol and dinosterol were also determined alongside IP25 to distinguish between two extreme scenarios, either ice-free or permanent ice conditions. PIP25 index values show a positive correlation with satellite-derived spring/summer sea-ice concentration. When calculating and interpreting the PIP25 proxy for reconstructions of Arctic sea-ice conditions, the uncertainties of the origin of the sterols need to be considered and thus more than one phytoplankton biomarker should be involved and the resulting PIP25 values should be compared. In the third study, Arctic Ocean sea-ice conditions during the Marine Isotope Stage (MIS) 3–1 were reconstructed by means of IP25 and phytoplankton-derived biomarkers data obtained from 7 sediment cores. The summer ice edge remained north of the Barents Sea even during extremely cold (i.e., Last Glacial Maximum (LGM)) as well as during warm periods (i.e., Bølling-Allerød). Spatial sea-ice conditions were reconstructed for the Central Arctic Ocean and adjacent areas by means of biomarker data from 16 sediment samples covering the LGM time-slice. The western Spitsbergen margin and northern Barents Sea margin were found to be productive regions. In contrast, the LGM high Arctic (>84°N) was covered by thick permanent sea ice throughout the year with rare break up. The spring/summer sea-ice margin significantly extended southwards during LGM in the marginal seas.