It is possible that the first autotroph used chemical energy rather than light. This could have been the main source of primary production after the initial inventory of abiotic organic material had been depleted. The electron acceptor most readily available for use by this first chemoautotroph would have been CO2. The most abundant electron donor may have been H2 that would have been outgassing from volcanoes at a rate estimated to be as large as 10(12) moles yr-1, as well as from photo-oxidation of Fe+2. We report here that certain methanogens will consume H2 down to partial pressures as low as 4 Pa (4 x 10(-5) atm) with CO2 as the sole carbon source at a rate of 0.7 ng H2 min-1 microgram-1 cell protein. The lower limit of pH2 for growth of methanogens can be understood on the basis that the pH2 needs to be high enough for one ATP to be synthesized per CO2 reduced. The pH2 values needed for growth measured here are consistent with those measured by Stevens and McKinley for growth of methanogens in deep basalt aquifers. H2-consuming autotrophs are likely to have had a profound effect on the chemistry of the early atmosphere and to have been a dominant sink for H2 on the early Earth after life began rather than escape from the Earth's atmosphere to space.