Information routing strategy is a hot issue in the study of complex network traffic dynamics. In this paper, we define a new gravitational centrality and introduce parameter α 1 to control the centrality of nodes. When α 1 are appropriate values, the gravitational centrality have the function of dividing the centrality of the same degree nodes. According to the gravitational centrality, we propose an effective gravitation path routing strategy in which the optimal paths between all pairs of nodes are chosen according to a cost function that incorporates gravitational centrality of nodes in paths. The purpose of this strategy is to improve network traffic capacity. The simulation results on the scale-free network show that our routing strategy is more effective than the efficient routing strategy proposed by Yan et al. [Phys. Rev. E, Stat. Phys. Plasmas Fluids Relat. Interdiscip. Top. 73, 046108 (2006)].INDEX TERMS Complex networks, scale-free network, congestion, traffic capacity.