Abstract: In order to analyze conformity in pedestrian's violation crossing behavior, structural characteristics and evolution rules of pedestrian network were studied by using the theory of network. First,some pedestrian violation crossing behavior was observed by the motion capture technology. Second based on the pedestrian's real crossing behavior,pedestrian network was built. Finally,pedestrian illegal crossing behavior trends were studied by analyzing the network degree under the conditions of different red light time and different gender distribution. It is concluded that when waiting time is longer, more pedestrians will join in the violation crossing once one pedestrian crossing on red signal occurs. In addition,male pedestrians are more likely to cross in conformity,and their behavior would attract other pedestrians committing violation.