Volatile organic compounds (VOCs) play a key role in air pollution of China. Among various sources, petrochemical industry is one of the important contributors, but its VOC emission estimate still exists a big uncertainty. Therefore, this study developed an inverse-dispersion calculation method (IDM), and applied it in a typical petrochemical plant that manufactures ethylene (170kt/year), polyethylene (100kt/year), and polypropylene (61kt/year), and determined the VOC emission amount for this complex industrial area source. Firstly, this study monitored VOC concentrations around this plant in April of 2017, and found that the VOCs at downwind receptors was obviously higher than the level at background, higher by 20.7ppb on average. This VOC increment was mainly contributed by ethylene (30.4%), propylene (17.8%), pentanes (16.4%), and butanes (13.4%), which is consistent with the knowledge of VOC components emitted from the manufacturing of ethylene, polyethylene, and polypropylene. Then, by using the inverse-dispersion calculation method (IDM), we determined the relationship coefficient γ between source emission rate and ambient concentration for each receptor of each test based on an assumed source emission rate, combined γ with the actual VOC concentrations measured in monitoring tests, and estimated the average VOCs emission of 666.0tons/year for this plant, including 18.1tons for ethane, 21.1tons for propane, 61.6tons for isobutane, 44.3tons for n-butane, 79.3tons for isopentane, 56.8tons for n-pentane, 115.4tons for ethylene, 102.5tons for propylene, 92.7tons for benzene, and 74.1tons for toluene. Our IDM estimate was in the same order with the traditional emission factor method estimate (916.4tons VOCs per year for this plant), and we believed the IDM can be applied to effectively estimate the VOCs emissions for those complicated industrial sources.