In order to evaluate the quality of data resulting from analyses of polychlorinated dibenzo-p-dioxins and dibenzofurans in soil, two Dutch laboratories have participated in a RIVM-coordinated interlaboratory comparison study, i.e. the Laboratory of Environmental and Toxicological Chemistry of the University of Amsterdam (UVA) and TAUW Infra Consult B.V. Subsamples of a sandy soil have been fortified with standards of 2,3,7,8-chlorine substituted PCDDs and PCDFs at three different levels (20-8600 ng/kg of soil). Each participant has determined the levels of the seventeen toxic congeners of PCDDs and PCDFs using the own methodology and reference standards. The subsamples have also been analysed by the RIVM. The accuracy and repeatability of the analysis has been examined by using the reported data from the complete analysis of three separate subsamples. Reported levels were generally higher than the actual levels of fortification. This was partly caused by background concentrations in the (blank) soil. Measurements of UVA and RIVM showed a mean relative standard deviation of approximately 5%. Those from TAUW were on average 14% at levels below 120 ng/kg of soil, whereas at higher levels RSDs were comparable with those from UVA and RUVM. It was concluded that comparable and reproducible analytical results are obtainable at levels above 200 ng/kg of soil. At lower concentrations, however, less accuracy and repeatability has to be considered.<br>