Road surface meteorological conditions are very important for traffic safety. According to the statistics, the ratio of traffic accidents that occur on icy, wet and dry road surface is about 4.2:1.6:1. A remote road surface meteorological condition sensor based on multi-wavelength light was developed in this paper which could identify the road surface condition including dry, wet, icy and snow, that may be vital for preventing traffic accidents. By using multi-wavelength optical remote sensing technology and near infrared light source, the diffuse reflectance spectrum of road surface can be detected. The four road surface states of dry, wet, ice and snow can be measured with non-contact method, and the thickness of pavement water film can also be measured. Such road sensors were used on G50 highway in Chongqing, China. Experiment results show that the system could meet the requirements well. ? 2019 SPIE.