The force chain network in aggregate blend indicates load transfer path and is the main carrier to resist external loads. The objective of this study is to analyze the force chain network characteristics of aggregate blend on the basis of complex network theory. The virtual aggregate blend models with three stone matrix asphalt (SMA) gradations and three dense asphalt concrete (AC) gradations were constructed using discrete element method (DEM). The force chain network topology structure of each aggregate blend was established according to the aggregate contact parameters extracted from DEM software and the introduced complex network theory. The characteristics of each force chain network were analyzed by the various network evaluation indicators. The results showed that there are two or more contacts between most particle pairs. The average coordination number and the average contact number of aggregate blend are not equal. The degree distribution of each gradation shows a good power-law distribution. The force chain network of aggregate blend is a scale-free network. Big particles have greater average strength and smaller average weight distribution difference compared with small particles. The reference lower limit has better clustering performance for SMA-16, and the reference upper limit has better clustering performance for AC-25. The average path length of SMA-16 is longer than that of AC-25, indicating that SMA-16 has better load transfer ability and skeleton properties.