Fine particles are significantly harmful to the human body and the atmospheric environment. However, the efficiency of current conventional dust removal devices for PM2.5 is low. Acoustic agglomeration is a promising pretreatment technology, which uses high-intensity sound wave to induce particles to agglomerate and improves the efficiency of conventional dust removal devices. In this study, an experimental setup of acoustic agglomeration was built and connected to an electrostatic precipitator (ESP) and a bag filter. The influence of acoustic agglomeration on dust removal efficiency was investigated. Results show that acoustic agglomeration has a significant positive effect on improving the dust removal efficiency of the ESP and bag filter. The removal efficiency increases with the increase in sound pressure level. The optimal frequencies or frequency ranges that are best for the combined effect are also obtained. When the combined system is operated at conditions of 1400 Hz and 148 dB, the particle mass concentration removal efficiency of the bag filter can be improved from 91.29% to 99.19% and that of the ESP can be improved from 89.05% to 99.28%. The influence of medium vibration caused by the high-intensity sound wave is also discussed. (C) 2015 Published by Elsevier B.V.