The air pollution transport has a great impact on regional air quality. Here, we adopted a complex network analysis approach to characterize the transport of fine particulate matter (PM_2.5) in the Yangtze River Delta (YRD) region with a resolution of 27 km × 27 km. We built a network to represent PM_2.5 transport in the YRD region and used a ranking algorithm to identify the important nodes in the network. We proposed two metrics to identify the important nodes: pollution output importance score (POIS) and pollution input importance score (PIIS). An improved PageRank algorithm, which includes air-mass-weight and pollution-transport-weight, was used to calculate each node’s POIS and PIIS, respectively. Both POIS and PIIS results showed that the prevailing transport pattern of PM_2.5 in the YRD region is North–South. Furthermore, POIS showed that the nodes in cities of Xuzhou, Changzhou, Huzhou and Jiaxing had higher contributions to PM_2.5 concentrations, while PIIS showed that the nodes in cities of Lishui, Huangshan, Anqing and south of Chizhou were the main receptor areas in the YRD region and were influenced most by the upstream pollution. This study established a complex network method for analyzing the transport pattern of regional air pollution.