Mechanisms influencing the ultimate outgassing rate of hydrogen from a metal chamber into a vacuum have been clarified by results from a reexamination using the pressure-rise method. The condition of the oxide surface of the metal plays a central role in determining the ultimate outgassing rate. Measurements were made for three materials, oxygen-free high-conductivity (pure) copper, chromium(0.6%)-copper(99.4%) alloy, and stainless steel 304, and are compared with previously published work by the throughput method. In the case of pure copper, the ultimate out, outgassing rate is dependent on in situ bakeout temperature. When the bakeout temperature is at least 250 degreesC, the oxide film on the copper surface is completely deoxidized and the ultimate outgassing rate is determined only by the concentration of hydrogen remaining in the bulk. As a result, this pure copper requires over a week in vacuum to attain its ultimate outgassing rate. In the case of chromium-copper alloy, however, the bulk content of hydrogen is mostly degassed during prebakeout at 400 degreesC, because the surface is not yet covered by a Cr2O3 film and the major copper-oxide film on the surface is easily deoxidized at this prebakeout temperature. During the prebakeout, Cr atoms in the alloy are out-diffused to the surface. The out-diffused Cr atoms on the surface are then oxidized to Cr2O3 when the chamber is subsequently exposed to atmospheric air. Because hydrogen permeation of a Cr2O3 film is over an order of magnitude lower than that for the bulk alloy, the film limits both out-diffusion of hydrogen from the bulk and hydrogen solution into the bulk through permeation when the surface is exposed to atmospheric air. Thus the outgassing rate of the CrCu chamber does not change with changes in bakeout temperature. The key to attaining ultralow outgassing rates in a pure copper chamber is in situ bakeout with extended, continued pumping; but the key for a CrCu alloy chamber is simply prebakeout. By this we can obtain an ultralow outgassing rate at the 10(-12)Pa m/s (hydrogen equivalent) level. On the other hand, in the case of stainless steel, the surface is immediately covered by a mixed Cr2O3+Fe2O3 film after the initial electropolishing step, because the content of Cr in stainless steel is over an order of magnitude higher than in the CrCu used. Therefore hydrogen in the bulk is not effectively degassed during 400 degreesC prebakeout and in situ bakeout. The outgassing rate from the stainless steel chamber was over an order of magnitude higher than from the pure copper and CrCu alloy chambers.