A comparative study of vacuum crushing methods for analyzing reactive gases in basaltic glasses shows that ball milling is an efficient means of releasing occluded gases but that blank production represents a potentially serious problem that must be accounted for when determining reactive-gas compositions. Production of H2 and CH4 in a stainless-steel ball mill in the absence of rock material increases with length of crushing time. However, test results presented here indicate that blank levels are reduced during the actual crushing process by the presence of rock powder, which may act as a cushion to reduce metal–metal contact. Crushing in copper tubes under a hydraulic press produces no blanks for these gases, but crushing efficiency and gas release are very low, and gas adsorption on rock powder becomes a significant problem. Experiments with methane adsorption on crushed basalt suggest that the loss of methane is a chemisorption process that is, for the most part, irreversible.Applying corrections for these effects, we find that H2/CH4 ratios (~3 – 30) in mid-ocean-ridge basalt glasses are similar to those seen in high-temperature mid-ocean-ridge hydrothermal fluids. These data, arguments based on the similarity of water/rock mass ratios calculated from basalt gas data, and the uniformity of methane/helium ratios in divers high-temperature mid-ocean-ridge hydrothermal fluids support the contention that dissolved CH4 and H2 in these fluids are predominantly derived from leaching of mid-ocean-ridge basalt.