Serpentinization of ultramafic rocks is widely recognized as a source of molecular hydrogen (H2) and methane (CH4) to support microbial activity, but the extent and rates of formation of these compounds in low-temperature, near-surface environments are poorly understood. Laboratory experiments were conducted to examine the production of H2 and CH4 during low-temperature reaction of water with ultramafic rocks and minerals. Experiments were performed by heating olivine or harzburgite with aqueous solutions at 90°C for up to 213 days in glass bottles sealed with butyl rubber stoppers. Although H2 and CH4 increased steadily throughout the experiments, the levels were very similar to those found in mineral-free controls, indicating that the rubber stoppers were the predominant source of these compounds. Levels of H2 above background were observed only during the first few days of reaction of harzburgite when CO2 was added to the headspace, with no detectable production of H2 or CH4 above background during further heating of the harzburgite or in experiments with other mineral reactants. Consequently, our results indicate that production of H2 and CH4 during low-temperature alteration of ultramafic rocks may be much more limited than some recent experimental studies have suggested. We also found no evidence to support a recent report suggesting that spinels in ultramafic rocks may stimulate H2 production. While secondary silicates were observed to precipitate during the experiments, formation of these deposits was dominated by Si released by dissolution of the glass bottles, and reaction of the primary silicate minerals appeared to be very limited. While use of glass bottles and rubber stoppers has become commonplace in experiments intended to study processes that occur during serpentinization of ultramafic rocks at low temperatures, the high levels of H2, CH4, and SiO2 released during heating indicate that these reactor materials are unsuitable for this purpose.