The concern about the future development of Arctic sea ice also prompts a gaining interest in past variations of the ice extent in the Arctic Ocean. The important question as to whether the palaeo variability in sea ice coverage in this climate-sensitive area can be firstly identified, and secondly, linked to climatic fluctuations thus motivates to look for tools that permit the reliable reconstruction of palaeo sea ice conditions. The recently established sea ice proxy IP25 is considered to be such a tool though a proper evaluation of the applicability of this biomarker hitherto has not been available. Within this thesis the occurrence of IP25 and its capability to display a previous sea ice coverage has been investigated by means of sediment cores from the Fram Strait - the only deep-water passage between the Arctic and Atlantic Ocean. Analyses of the biomarker composition of surface sediments revealed that the abundance of IP25 in these samples mirrors the recent, as derived from satellite and model data, sea ice distribution considerably well. It is found that the ice cover does not necessarily hamper but instead even may promote the phytoplankton productivity. This finding led to the establishment of a phytoplankton-IP25 (PIP25) index that facilitates the identification of different sea surface conditions (no ice, less/variable ice, marginal ice, perennial ice cover). Correlation analyses of these PIP25-based sea ice estimates with observed (and modelled) sea ice concentrations substantiate that this combinatory biomarker approach permits a more precise assessment of sea ice coverage than IP25 alone and could be used for quantitative (palaeo) sea ice reconstructions. Further, the Holocene variability in IP25 and phytoplankton biomarker contents of sediment cores from the continental margin of West Spitsbergen and East Greenland was studied. Throughout the Middle to Late Holocene, the sea ice cover increased in eastern Fram Strait, whilst the sea ice export along the East Greenland coast continued and remained relatively unaffected by short-term variations in the oceanic circulation. Such fluctuations led to rapid sea ice oscillations along the West Spitsbergen margin during the Neoglacial that can be linked to abrupt changes in the advection of Atlantic water (possibly also due to shifts in the atmospheric NAO-like forcing). In a third study, which also represents the first long-term application of IP25, extreme variations in the sea ice cover during the last glacial-interglacial transition were reconstructed for northern Fram Strait. Permanent ice coverage is assumed for the Last Glacial Maximum such that the productivity of ice algae and phytoplankton at the sea surface was inhibited. Abrupt short-term cooling and warming events (e.g. the Intra-Allerød cold period and the Bølling warm phase) were correlated convincingly with sea ice advances and retreats, respectively. Within this thesis it is demonstrated that the approach to combine IP25 with phytoplankton-derived biomarkers occurs to be a promising and reliable method for palaeoceanographic sea ice reconstructions and which increases the significance of IP25. The studies provide information about the usage and interpretation of IP25 and they give new insight into the evolution of the sea ice conditions in Fram Strait over the past 30,000 years. Future studies on the spatial and temporal distribution of IP25 will likely extend the approach to use this proxy for quantitative sea ice reconstructions and also to elucidate sea ice (and climate) conditions of ancient times.