Stainless steel exposed to gaseous tritium characteristically shows a firmly trapped fraction of tritium in the surface layer, which is not fully removable by water at ambient temperature. Prolonged thermal treatment of tritium-loaded specimens at <443 K causes substantial depletion of the bulk but almost no depletion of the surface layer. For complete removal of hydrogen isotopes from the bulk and the surface, temperatures exceeding 573 K are necessary. Upon chemical etching virtually all tritium trapped in the surface layer appears in the etching solution as tritiated water. Following removal of the layer by chemical etching, the tritium-rich layer reappears after months of aging at ambient temperature with nearly the original tritium activity. Comparison of chronic tritium release rates into liquid water before and after etching reveals that the surface layer only marginally influences the rate. X-ray photoelectron spectroscopy provides evidence that during prolonged aging the surface layer continues to grow while at the same time trapping a fraction of bulk tritium released at ambient temperature. Experimental results suggest different mechanisms of hydrogen uptake and release by the bulk and surface layers. Inference of tritium activity in the bulk of aged or heat-exposed stainless steel material from surface activity measurements may depart significantly from reality.