[en] A quality control study was organized on two air filters artificially loaded with urban dust. One dust material was collected from the air intake ducts of the Vienna General Hospital, the second material was collected from the main ventilation shaft of a road tunnel in Prague. 49 laboratories from 40 countries participated in the study, providing 53 sets of results. The following analytical methods were used: neutron activation analysis (NAA), inductive coupled plasma atomic emission spectrometry (ICP-AES), inductively coupled plasma mass spectrometry (ICP-MS), particle induced X-ray emission spectrometry (PDCE), atomic absorption spectrometry (AAS), and X-ray fluorescence spectrometry (XRF). The results were evaluated following the standard LEA procedures for (1) an interlaboratory comparison exercise and (2) a proficiency test using established target values. For the air filter sample V, 754 laboratory mean values for 38 elements were reported. More than 4 mean values were obtained for 26 measurands and these were statistically evaluated. Only 10 % 71 values) of the results were detected as outliers by the statistical data evaluation. The results and statistical parameters are listed in the appendices I and II. The results were also evaluated as a proficiency test using established target values. Precision and accuracy criteria were applied to 15 elements. 534 laboratory mean values were reported for those elements, 114 values (20%) did not pass the set proficiency test criteria. The individual laboratory results are listed in appendix IV. For the air filter sample P, 694 laboratory mean values for 38 elements were reported. More than 4 mean values were obtained for 26 measurands and these were statistically evaluated. Only 9% (60 values) of the results were detected as outliers by the statistical data evaluation. The results and statistical parameters are listed in the appendices I and II. The results were also evaluated as a proficiency test using target values from the literature. Precision and accuracy criteria were applied to 10 elements. 350 laboratory mean values were reported for these elements, 75 values (21%) did not pass the set proficiency test criteria. The individual laboratory results are listed in appendix IV. In conclusion, the results from this quality control study revealed that the determination of trace elements in dust material loaded on filters is a difficult task for analytical laboratories. The results indicate the urgent need for a real matrix standard reference material for checking and improving the analytical performance in laboratories analysing routinely air filters (author)