Data and supporting documentation were compiled and analyzed for 26 cases of gas grab samples taken during waste-tank core sampling activities between September 1, 1995 and December 31, 1997. These cases were tested against specific criteria to reduce uncertainties associated with in-tank sampling location and conditions. Of the 26 possible cases, 16 qualified as drill-string grab samples most likely to represent recently released waste gases. The data from these 16 ``confirmed`` cases were adjusted to remove non-waste gas contributions from core-sampling activities (argon or nitrogen purge), the atmospheric background, and laboratory sampler preparation (helium). The procedure for subtracting atmospheric, laboratory, and argon purge gases was unambiguous. No reliable method for determining the exact amount of nitrogen purge gas was established. Thus, the final set of ``Adjusted`` drill string gas data for the 6 nitrogen-purged cases had a greater degree of uncertainty than the final results for the 10 argon-purged cases. Including the appropriate amounts of uncertainty, this final set of data was added to the set of high-quality results from the Retained Gas Sampler (RGS), and good agreement was found for the N{sub 2}, H{sub 2}, and N{sub 2}O mole fractions sampled from common tanks. These results indicate that undermore » favorable sampling conditions, Drill-String (DS) grab samples can provide reasonably accurate information about the dominant species of released gas. One conclusion from this set of total gas data is that the distribution of the H{sub 2} mole fractions is bimodal in shape, with an upper bound of 78%.« less