Abstract:Knowledge transfer within multinational corporation is great significance to enhance the multinational competitive advantage. From the perspective of network structure and network nodes, four factors including network size, network density, transfer motivation and absorptive capacity affecting knowledge transfer within multinational corporation internal network are analyzed. A theoretical model and a simulation model based on multi-agent, which refer to knowledge transfer within multinational corporation internal network, are built. Simulation experiments are done with NetLogo and simulation results are analyzed by MANOVA. Main effect and interaction are verified. Some conclusions are draw and some advice to Chinese multinational corporations which implement for foreign direct investment and multinational business are given.