In the context of global element recycling at subduction zones, altered oceanic mantle probably plays a major role for light elements. Despite this, its inventory of these elements has hardly been explored. We performed a systematic study on the abundance, partitioning and mobility of B, Li, Be and the B and Li isotope fractionation in rock-forming minerals from altered oceanic mantle. In order to constrain the enrichment of oceanic mantle in light elements by hydrothermal and seafloor alteration and thus probably the major input of light elements in subduction zones we studied samples from the mid-Atlantic slow-spreading ridge (ODP collection) and non-metamorphic ophiolites from Greece and the Swiss Alps. The most important methods employed include field and petrographic work, electron microprobe analyses, secondary ion mass spectrometry (SIMS) analysis for B, Li and Be concentrations as well as multicollector inductively coupled plasma mass spectrometry (ICP MS) and thermal ionisation mass spectrometry (TIMS) for B, Li and Sr isotope analysis. Our results show that oceanic mantle at slow-spreading ridges becomes enriched in B and in 11B through the formation of serpentine, a hydrous phyllosilicate, during complex seawater-rock interaction. It attains high B concentrations and extremely heavy ?11B. At the same time, Li and preferentially 7Li are leached from the oceanic mantle, leaving it with very low Li concentrations (mainly hosted by clinopyroxene) and extremely light ?7Li. Be is hardly present at detectable levels in oceanic mantle. The implications of our results are that mantle formed at slow-spreading ridges can potentially carry high B concentrations and extremely heavy ?11B into subductions zones but that it plays virtually no role for the Li budget. The type of mantle we have studied dominated the Jurassic / Cretaceous Tethys ocean and was in large parts subducted in the Eocene. In the modern oceans, it probably makes up 50-70% of the mantle.