One of the main problems in the VANET(vehicular ad-hoc network)routing algorithms is how to establish the stable routes.The link duration in these networks is often very short because of the frequent changes in the network topology.Short link duration reduce the network efficiency.Different speeds of the vehicles and choosing different directions by the vehicles in the junctions are the two reasons that lead to link breakage and a reduction in link duration.Several routing protocols have been proposed for VANET in order to improve the link duration,while none of them avoids the link breakages caused by the second reason.In this paper,a new method for routing algorithms is proposed based on the vehicles trips history.Here,each vehicle has a profile containing its movement patterns extracted from its trips history.The next direction which each vehicle may choose at the next junction is predicted using this profile and is sent to other vehicles.Afterward each vehicle selects a node the future direction of which is the same as its predicted direction.Our case study indicates that applying our proposed method to ROMSGP(receive on most stable group-path)routing protocol reduces the links breakages and increases the link duration time. One of the main problems in the VANET (vehicular ad-hoc network) routing algorithms is how to establish the stable routes. The link duration in these networks is often very short because of the frequent changes in the network topology.Short link duration reduce the network efficiency. Different speeds of the vehicles and choosing different directions by the vehicles in the junctions are the two reasons that lead to link breakage and a reduction in link duration. Routing protocols have been proposed for VANET in order to improve the link duration, while none of them avoids the link breakages caused by the second reason. this paper, a new method for routing algorithms is proposed based on the vehicles trips history. Here, each vehicle has a profile containing its movement patterns extracted from its trips history. The next direction which each vehicle may choose at the next junction is predicted using this profile and is sent to other vehicles. Afterward each vehicle selects a node the fu ture direction of which is the same as its as predicted direction. Our case study that apply our proposed method to ROMSGP (receive on most stable group-path) routing protocol reduces the links breakages and increases the link duration time.