An apparatus and technique were developed to measure the solubility of tritium in high diffusivity metals at tritium pressures up to 136 MPa (20,000 psia) and temperatures up to 700 K. The experiments described in this paper took advantage of the low detection limits for tritium and helium-3 to determine the solubilities of high pressure tritium in copper and gold at temperatures as low as 473 K, where solubilities are below the limits of detection for hydrogen or deuterium. Samples were exposed to high pressure tritium at an elevated temperature long enough to reach equilibrium, then cooled within seconds by dropping them into a cold (77 K) section of the apparatus, thereby immobilizing the dissolved tritium. Solubilities were then determined by acid dissolution/liquid scintillation counting or, alternatively, by vacuum fusion/helium-3 analysis. For both copper and gold, surface effects were found to be extremely important because (1) they greatly increased the time required for the samples to equilibrate with the tritium overpressure and (2) there was more tritium on and near the surface than was contained in the bulk of a sample. In the absence of trapping, the solubilities determined at high pressures at temperatures between 473 K and 673 K agreed well with extrapolations of solubilities measured at 0.1 MPa hydrogen and temperatures >873 K. Gold annealed at 1273 K in air exhibited much higher apparent solubilities than samples annealed at 873 K in air or vacuum, an effect tentatively attributed to trapping by oxygen which diffused into the sample during the high temperature annealing treatment.