Keep-Right-Except-To-Pass rule is accepted by most countries.Firstly, we build differrrent models to evaluate the performance of the rule. In macroscopic aspect,we build the probabilistic model of demand for overtaking after fitting time headway distribution,proving that it is unable to give consideration to both traffic flow and safety.Therefore,we create a comprehensive evaluation function to reflect the performance of the rule.It indicates that in the case of light traffic, with traffic volume increases, the traffic condition becomes better;However,in the case of heavy traffic, the conclusion is that with traffic volume increases, the traffic condition becomes worse. As to be microcosmic, we established a single overtaking model based on its physical process. It defines that the Minimum/Maximum Road Capacity is the minimum/maximum allowed number of overtaking occurs simultaneously. Safety increases with the Maximum Road Capacity and traffic volume enlarges with the Minimum Road Capacity. The conclusion was drawn from the model test of real situation of freeway. The traffic volume increases as the maximum speed limit decreases, while the safety stays stable; Safety gets worsened when the minimum speed limit was increased, while the traffic volume stays almost still. We create a new overtaking rule. Taking highways with three lanes for instance, the limitation of the speed should increase from the right to the left lanes. Overtaking is prohibited on the lane with highest speed. As for cars on other two lanes, if they want to pass another vehicle, they should move to the adjacent left lane. We conduct the cellular automaton to simulate the cases under the new rule,keep-right-except-to-pass rule and without no rule. We conclude that the new rule we suggest generally perform better than keep-right-except-to-pass rule. As for left-driving countries, we can simply change the orientation in the single overtaking model. we design an intelligent transport system based on road network and intelligent vehicles. In addition, it is highlighted that our design take the drivers’ satisfaction into consideration to make it a humanization system. Furthermore, we implement computer simulation and comprehensive evaluation on this system.