The lateral position of a passing vehicle crossing the Orthotropic Steel Deck (OSD) is crucial for both OSD bridge fatigue analysis and the Bridge Weigh-In-Motion (BWIM) system. In regard to the local salient effect of OSD bridges, an efficient method is proposed in this work that only uses the collected signals of sensors applied to the BWIM system to identify the lateral position of the passing vehicles. This method establishes the finite element model of OSDs, which can be verified during the calibration test in the BWIM system. The influence lines of girders are then extracted to form the error function between the theory responses and measurements. To verify the accuracy and feasibility of the proposed method, the Ls-dyna-based vehicle and bridge interaction model is first built and analyzed.The effect of vehicle conditions (including the vehicle speed, the number of vehicle axles, and the different wheel lateral positions on the bridge) and wheel-to-bridge contact area on the recognition accuracy is investigated. The simulation analysis shows that the proposed method is not sensitive to vehicle conditions. For different vehicle speeds, the number of axles, the lateral position of the wheels, and the max average errors of the recognition accuracy are 9 mm, 4.6 mm, and 12.5 mm, respectively. In addition, it reveals that the suitable wheel-to-bridge contact area should be 200 mm. Then, a field bridge test is carried out in this paper to verify the effectiveness of this method. The field test results show that the proposed method has high accuracy in identifying the lateral position of the vehicle and has a wide range of applications.