We developed a simple system using tritium tracer and thermal desorption techniques to measure the tritium adsorption and/or absorption on/in a material having typical surface conditions: namely, not cleaned surface. The tritium counting devices used were a 2π counter and conventional proportional counter. With this system, the amounts of ad/absorption could be measured without exposing the samples to air after exposing them to tritium gas. The overall efficiency (F) of the 2π counter was described as F = exp(−2.64h), where h is the distance from the sample to the detector. Ad/absorption measurements were carried out for several materials used for fabricating conventional vacuum systems. The results were, in the order of decreasing amounts of ad/absorption, as [fiber reinforced plastics(FRP)] > [nickel(Ni), molybdenum disulfide(MoS2)] > [stainless steel (SS304), iron(Fe), aluminum alloy(A2219)] > [boron nitride(h-BN), silicon carbide(SiC), SS304 passivated by anodic oxidation layers(ASS) and that by boron nitride segregation layers(BSS)]. The relative amounts were about 100 for Ni and 0.1 for ASS and BSS, being normalized to Fe = 1. It was found that the passivation of SS304 with anodic oxidation layers and/or BN segregation layers should be quite valid to decresase the tritium inventory on/in the material walls of tritium handling systems. In addition, it was estimated that this system would be capable of detecting the tritium adsorption of the order of 10−6 in the surface coverage.