In order to reveal the industrial agglomeration pattern, multi-scale characteristics, and driving mechanism of the air-rail dual-hub regions, this study takes Guangzhou as an example to extract the universal patterns of this region type. Based on POI big data of 9 different types of industries, this study uses kernel density analysis, average nearest neighbor analysis, Ripley’s K function analysis, and spatial autocorrelation analysis to explore the agglomeration characteristics of various industries in the air-rail dual-hub regions. The results show that the industries have a “core-edge” spatial distribution pattern: high-value areas are mainly concentrated around Guangzhoubei Railway Station and Huadu Central Business District, and extend in a belt and bead chain along rail transit and main roads. The overall industries show significant agglomeration, and the intensity is gradient. The consumption-oriented industries are stronger (food and beverage service 0.306, accommodation service 0.331, shopping service 0.338); Culture and education services, enterprises, and sports and recreation services are relatively weak (0.433-0.489); The agglomeration is hierarchical in scale, and the dominant scale is between 1.92 km and 4.48 km, forming a multi-scale structure with the coexistence of “node attachment-mesoscale composite service circle-regionalized cluster superposition”; Industrial POI density is positively correlated with road network density, night light index and population density as a whole, showing the composite driving characteristics of “population support-traffic guidance-vitality enhancement”. The results show that the hub effect of the air-rail dual-hub regions has expanded from the agglomeration around the station to the multi-type composite organization mode of “circle-corridor-group”. The study provides a basis for the guidance of spatial planning in different types and scales of the air-rail dual-hub regions.