Abstract: Airborne particulate matter (PM), especially PM_2.5, can be easily adsorbed by human respiratory system. Their roles in carrying pathogens for spreading epidemic diseases has attracted great concern. Herein, we developed a novel gelatin filter-based and culture-independent method for investigation of the microbial diversity in PM samples during a haze episode in Tianjin, China. This method involves particle capture by gelatin filters, filter dissolution for DNA extraction, and high-throughput sequencing for analysis of the microbial diversity. A total of 584 operational taxonomic units (OTUs) of bacteria and 370 OTUs of fungi at the genus level were identified during hazy days. The results showed that both bacterial and fungal diversities could be evaluated by this method. This study provides a convenient strategy for investigation of microbial biodiversity in haze, facilitating accurate evaluation of airborne epidemic diseases.