The overall goal of the IPY projects WARMPAST and SciencePub is to advance our knowledge of climate warming in the Arctic, by studying past climate change. In order to investigate natural rapid ocean changes on longer time scales it is crucial to obtain reliable quantitative proxy data. Earlier reconstructions of sea surface temperatures (SST) below 5°C are hampered by incomplete training sets, sample quality and unsuitable sample preparation techniques. The aim of this study is to obtain surface samples with undisturbed sediment-water interface and correlate modern planktonic foraminifera with in situ measurements of environmental parameters e.g. sea-ice distribution, temperature and salinity. Other hampering factors have been to find suitable statistical methods or that the reconstructed SST do not always reflect the upper most water masses as depth habitats of planktonic foraminifera are variable. The studied size fractions of planktonic foraminifera are also a critical factor. Many studies have investigated the fraction larger than 125 and 150 μm. This may reduce the faunal diversity to near monospecific assemblages and much vital fauna information is left out of the SST reconstructions. Surface sediment samples from the Fram Strait, Greenland and Barents Seas were retrieved using multicorer/boxcorer. The samples were wet sieved at the size fraction 0.1 – 1.0 mm. The results show an overall dominance by N. pachyderma (sinistral) in surface waters dominated by Arctic and Polar waters masses. In surface waters dominated by Atlantic water masses, other species like T. quinqueloba and G. uvula increase. T. quinqueloba is particular frequent at oceanographic fronts.