Adapted prophylactic measures combined with the use of resistant cultivars is, up to now, the most effective way to reduce the incidence of Ralstonia solanacearum, the causal agent of bacterial wilt. In order to optimize the efficiency of prophylactic measures, powerful identification and detection tools of the bacterium in any potential inoculum source (plant, seed, water, soil) are required. However, the commonly used methods such as isolation on semi-selective medium (2, 7, 10), serological methods such as ELISA or immunofluorescence (6, 14), or pathogenicity tests on host plants (4, 9) for the diagnosis of bacterial wilt are often inadequate in terms of specificity, sensitivity or response time, especially for detecting the bacterium in soil. DNA amplification offers many advantages over the above-mentioned techniques such as increased specificity, sensitivity and response time. Nevertheless, the PCR method has not yet become the diagnostic tool of choice for laboratories, mainly because of the inhibition of the amplification reaction by compounds contained in crude bacterial extracts which give false negative results or low detection sensitivity. Although a wide range of inhibiting substances have been reported, the identity and mode of action of most of them remain unclear (17). The aim of this study was to compare several procedures to overcome PCR inhibition problems and to propose protocols ensuring a reliable PCR detection of low levels of R. solanacearum populations in natural settings.