A collaborative study has been performed over a period of 3 years to develop a suitable method for monitoring the mutagenicity of airborne particulate matter. The study was organized with 8 laboratories and performed in the following steps: (1) selection of a suitable technique for each process involved in the mutagenicity monitoring, (2) developing a tentative protocol by combining systematically the selected techniques, (3) evaluation of the protocol by intra- and inter-laboratory studies, (4) modification of the protocol according to the evaluation, and (5) evaluation of the modified protocol by conducting an interlaboratory study. We found a suitable method for mutagenicity monitoring of particles in the atmosphere. Airborne particles were sampled with a high-volume sampler, the samples were stored at -80 degrees C, extracted by sonication using dichloromethane, solvent-exchanged, and assayed by the preincubation method using Salmonella typhimurium TA98 and TA100. The observed mutagenic activity was normalized with that of an internal standard. Round robin tests revealed that the method resulted in excellent reproducibility. The coefficient of variation for mutagenic activities of airborne particulate samples collected in various districts of Japan were in the range of 14.7 +/- 6.6% to 19.6 +/- 4.0% for strains TA98 and TA100 with and without metabolic activation. We also found that the plate incorporation method was equivalent to the preincubation method for airborne particulate extracts.