With the development of the economy and the rapid increase in traffic volume, an overload phenomenon often occurs. This paper studied a vehicle load identification technique based on orthotropic bridge deck stress monitoring data. The strain responses on the lower edge of multiple U-ribs were collected under vehicles crossed the deck. Firstly, an index based on the cross-correlation function of strain response between different measurement points on the same U-rib was used to evaluate vehicle speed. Secondly, a cosine similarity index was proposed to locate the transverse position of the vehicle. Finally, the unknown vehicle load was identified on the basis of a calibrated strain response area matrix. The effectiveness and anti-noise performance of the proposed method were verified using numerical simulation. An experimental model was designed and some strain gauges were installed to measure the strain response, and the test was carried out to further verify the algorithm’s performance. Numerical and experimental results show that the proposed method could effectively identify the vehicle load with good anti-noise performance. Moreover, a calibration space was provided to guide practical engineering applications. The proposed method does not damage bridge decks, does not affect traffic, and is economical.