In recent years, a novel proxy for the past occurrence of Arctic sea ice has been proposed that is based on the variable marine sedimentary abundance of an organic geochemical lipid derived from sea ice diatoms in the spring. This lipid, termed IP25 (Ice Proxy with 25 carbon atoms), is a highly branched isoprenoid mono-unsaturated alkene that appears to be sufficiently stable in sediments to permit meaningful palaeo sea ice reconstructions to be carried out over short- to long-term timescales. Since the first proposed use of IP25 as a proxy for palaeo sea ice by Belt etal. (2007), a number of laboratories have measured this biomarker in Arctic sediments and it is anticipated that research activity in this area will increase further in the future. The content of this review is divided into a number of sections. Firstly, we describe the scientific basis for the IP25 proxy and its initial discovery in Arctic sea ice, sedimenting particles and sediments. Secondly, we summarise the relatively few studies that have, to date, concentrated on examining the factors that influence the production and fate of IP25 and we identify some areas of future research that need to be addressed in order to improve our understanding of IP25 data obtained from sedimentary analyses. What is clear at this stage, however, it that the presence of IP25 in Arctic marine sediments appears to represent a proxy measure of past seasonal seaice rather than permanent or multi-year ice conditions. Thirdly, we highlight the importance of rigorous analytical identification and quantification of IP25, especially if measurements of this biomarker are going to be used for quantitative sea ice reconstructions, rather than qualitative analyses alone (presence/absence). Fourthly, we review some recent attempts to make the interpretations of IP25 biomarker data more detailed and quantitative by combining sedimentary abundances with those of phytoplankton- and other sea ice-derived biomarkers. Thus, the bases for the so-called PIP25 and DIP25 indices are described, together with an overview of potential limitations, concluding that investigations into the use of these indices needs further research before their full potential can be realised. In the final section, we provide a summary of IP25-based palaeo sea ice reconstruction case studies performed to date. These case studies cover different Arctic regions and timescales spanning decades to tens of thousands of years. © 2012 Elsevier Ltd.