The deposition of the airborne microorganisms onto cultural heritage is associated closely with the subsequent biodeterioration. In this study, a systematic investigation was carried out to detect the seasonal variation and diversity of airborne fungal concentration at the World Cultural Heritage Site Maijishan Grottoes in western China. A bio-aerosol sampler was deployed to collect samples over four seasons in 2016. The culturable airborne fungi were isolated, purified and then identified with the extraction of genomic DNA, PCR amplification of ITS rRNA region, sequencing, and phylogenetic analysis. The concentrations of culturable fungi ranged from 216 to 1389 CFU/m³, which varied seasonally with significant differences among the sampling sites. Fifteen different fungal genera were confirmed, among them, Cladosporium was the most predominant fungal genus, followed by Penicillium. The fungal community structure and their relationship with environmental factors were also delineated. The spatial–temporal differences of airborne fungi at Maijishan Grottoes were mainly due to height, rainfall, relative humidity, and temperature. The dominant genera Cladosporium and Penicillium may pose potential threats to the ancient painted sculptures and murals, and monitoring of the airborne fungi at such a heritage site could provide supporting data for the pre-warning and control of fungal outbreaks inside the caves for better management.