Successful applications of air sampling for the quantification of exposure to indoor allergens have been reported, but its efficiency is still controversial. We evaluated whether the low-flow, long-term air sampling in normal domestic activity conditions can quantify the exposure of house dust mites (HDM) and cockroaches (CR) allergens or not. Airborne Der f 1 and Bla g 1 were captured with a personal air sampler in 25 bedrooms during normal domestic activity. Quantification of the major allergens in the reservoir dust and the extraction of the air sampler filters were done with two-site ELISA kits. Airborne Der f 1 was measured above the threshold level of detection in 15 houses (60%). Detection rate of airborne Der f 1 was significantly higher in those houses where D. farinae was microscopically found in the reservoir dusts (76.5% vs. 25%, chi 2 = 6.0, p = 0.014). Airborne Der f 1 was more frequently detected in the houses with higher Der f 1 (> or = 10 micrograms/g dust) in bedding reservoir dust than the other group (91% vs. 35.7%, chi 2 = 7.819, p = 0.005), and the median value of airborne Der f 1 was also significantly higher in that group (14.0 pg/m3 vs. below detection limit, p = 0.002). Airborne Der f 1 was significantly correlated with Der f 1 in bedding reservoir dust (r = 0.591, p < 0.01). Airborne Bla g 1 was measured with ELISA in 16 houses (64%), and it was more frequently detected in the houses where the CRs were captured by adhesive traps (91% vs. 57%, chi 2 = 3,484, p = 0.06). The median concentration of Bla g 1 in the filter was also higher in the houses with captured CRs (0.12 vs. 0.05 mU/m3, p = 0.06), but the level of Bla g 1 did not correlate with that of the bedding dusts or the floor dusts of kitchen. These results suggested that airborne HDM or CR allergens could be measured by low-flow, long-term air sampling, and that it might be one of appropriate modalities for evaluating personal exposure to HDM and CR allergens.