Large amounts of methane are produced in marine sediments but are then consumed before contacting aerobic waters or the atmosphere. Although no organism that can consume methane anaerobically has ever been isolated, biogeochemical evidence indicates that the overall process involves a transfer of electrons from methane to sulphate and is probably mediated by several organisms, including a methanogen (operating in reverse) and a sulphate-reducer (using an unknown intermediate substrate). Here we describe studies of sediments related to a decomposing methane hydrate. These provide strong evidence that methane is being consumed by archaebacteria that are phylogenetically distinct from known methanogens. Specifically, lipid biomarkers that are commonly characteristic of archaea are so strongly depleted in carbon-13 that methane must be the carbon source, rather than the metabolic product, for the organisms that have produced them. Parallel gene surveys of small-subunit ribosomal RNA (16S rRNA) indicate the predominance of a new archael group which is peripherally related to the methanogenic orders Methanomicrobiales and Methanosarcinales.