Abstract: To solve the problem of commercial vehicle mass estimation in mountain area, a mass estimation algorithm was proposed for vehicle. Firstly, a double layer structure algorithm was designed based on a vehicle longitudinal dynamics model for commercial vehicle mass estimation. The upper layer of the algorithm was arranged for road grade estimation based on Tilt sensor, while the lower level of the algorithm was for the vehicle mass estimation based on the recursive least squares algorithm with time varying forgetting factor. Then, the influence of vehicle suspension on the upper layer algorithm was analyzed with TruckSim software, and the vehicle experiment was carried out. The experimental results show that the proposed algorithm can estimate road slope and vehicle mass effectively. The algorithm can improve the estimation accurate and the convergence speed. The mean square error of modified vehicle mass can be reduced from 209.97 kg to 117.43 kg.