Road test simulation on test rig is widely used in the automobile industry to shorten the development circles. However, there is still room for further improving the time cost of current road simulation test. This paper described a new method considering both the damage error and the runtime of the test on a multi-axial test rig. First, the fatigue editing technique is applied to cut the small load in road data to reduce the runtime initially. The edited road load data could be reproduced on a multi-axial test rig successfully. Second, the rainflow matrices of strains on different proving ground roads are established and transformed into damage matrices based on the S-N curve and Miner rules using a reduction method. A standard simulation test for vehicle reliability procedure is established according to the proving ground schedule as a target to be accelerated. Third, with the runtime and the damage matrix of each road, a multi-objective optimal model is built based on the principle of equivalent damage. In the objective function, there are two optimization objectives. One is minimizing the runtime on the test rig, and the other is minimizing the error of damage matrix between the target and the results. Finally, the Non-Dominated Sorting Genetic Algorithm II (NSGA-II) is used to solve the problem. The Pareto curve of the two optimization objectives (errors vs. runtime) showed that a high accelerate ratio corresponds to a relatively high damage equivalent error.