Abstract: Oolong tea is one of the six major Chinese traditional tea categories, due to its naturally fruity and floral aroma. Unlike green and black tea, the quality of oolong tea is originated from the processing (named green-making), which also contributes to the semi-fermentation degree. The green-making includes turnover and indoor withering. The former refers to shaking the solar-withered tea leaves several times, while the latter is to pause for hours after turnover. The two steps are repeated alternately during the manufacturing process of oolong tea. Therefore, all process parameters can pose some influences on the aroma of oolong tea, such as the mode of green-making, the intensity of turnover, indoor withering, and the thickness of spread tea leaves. In addition, the environmental factors of green-making also play an important role in the formation of oolong tea quality, including environmental temperature, humidity, and atmosphere. The airflow is the carrier of temperature and humidity, particularly the indispensable determiner of oolong tea quality. However, it is still lacking in the effects of airflow on the oolong tea green-making during tea processing. This study aims to explore the effect of the airflow factor on the aroma quality of oolong tea during the process of green-making. The test materials were selected as fresh tea leaves (CK), normoxic tea leaves (Ta), hypoxic tea leaves (Tb), and their associated raw tea a (Ma) and raw tea b (Mb). The volatile components of oolong tea mediated by different airflow factors were analyzed by multivariate statistical analysis after the semi-quantitative Reverse Transcription and Polymerase Chain Reaction (RT-PCR), headspace solid-phase microextraction (HS-SPME), and Gas Chromatography Time-of-Flight Mass Spectrometry (GC-TOF MS). The result showed that both green-making rollers (named subroller a and subroller b) fully met the conditions of normoxia and hypoxia. The relative expression levels of hypoxic stress response factors CsHIG4 gene and CsHIG6 gene in Tb were up-regulated founded on the brightness of electrophoresis bands compared with the Ta and CK. 205 volatile components were identified in the raw tea. 33 kinds of shared differential metabolites were found among the CK, Ta and Tb, most of which were alcohol (8 kinds) and alkene (7 kinds). Partial Least Squares (PLS) analysis showed that there were 24 characteristic aroma components, whose Variable Important Projection (VIP) values were more than 1.0. Among them, the VIP value of indole was the largest (4.75), followed by α-farnesene (3.06), ocimene mixture of isomers (2.77), and trans-nerolidol (2.75). The analysis of variance showed that there were five characteristic aroma components in the Mb, whereas, only two in the Ma. The sensory evaluation of raw tea indicated that the aroma of Ma was clean, lasting and refreshing, indicating a distinct flowery scent, achieved the final score 90.67. More importantly, the final score was achieved in 90. By contrast, the aroma of the Mb was dull, less pleasant, lack of long and clean, particularly with the floral and characteristics variety fragrance. The contents of five volatiles in the Ma were extremely significantly higher than those in the Mb (P<0.05), including the trans-nerolidol, 2,3-dihydro-3,5-dihydroxy-6-methyl-4(H)-pyran-4-one, isophytol, 2-ethylhexyl butyl sulfite, 6-azabicyclo 3.2.1 octane. It infers that these five substances can be the landmark metabolites to form better quality oolong tea under the normoxic turn-over condition. This finding can provide a strong reference to better understand the effect of airflow factors on the metabolism of aroma substances during the green-making process. A theoretical foundation was also laid to explore the formation mechanism of aroma quality during the green-making process of oolong tea mediated by environmental factors.