Until recently all archaebacteria isolated conformed to one of three basic phenotypes: they were either methanogens, extreme halophiles, or ('sulphur-dependent') extreme thermophiles1. However, a novel phenotype, that fits none of these categories, has recently been described2. The organism, strain VC-16 (tentatively called "Archaeoglobus fulgidus") reduces sulphate—the only archaebacterium so far known to do so—and makes very small quantities of methane, although it lacks some of the cofactors normally associated with methanogenesis2. These characteristics suggest that strain VC-16 might represent a transition form between an anaerobic thermophilic sulphur-based type of metabolism (which seems to be the ancestral metabolism for archaebacteria3,4) and methanogenesis (which somehow then derives from it). We here show that the lineage represented by strain VC-16 arises from the archaebacterial tree precisely where such an interpretation would predict that it would, between the Methanococcus lineage (which is the deepest of the methanogen branchings) and that of Thermococcus (the deepest of all branchings on the methanogen side of the tree).