Road surface temperature forecast is a key component of winter maintenance strategy in many developed countries. Numerical tools exist to help road managers to organize services and consequently to trigger de-icing operations. Forecasting strategies have been commonplace since the 1980s, and often based on numerical models. Traffic is one of the influencing parameters, specifically in urban areas. This work was undertaken to evaluate to which extent an accurate description of traffic might improve numerical model dedicated to road surface temperature forecasting. Two sets of experiments were run to detect and to quantify traffic effects on RST. First one consisted in driving above an infrared radiometer, a pyrgeometer and other atmospheric probes to measure the radiative contribution of a passing vehicle at various speeds. In the second set, an infrared camera was installed on a vehicle in an urban traffic flow. This camera was mounted on the roof and focused the pavement right behind the vehicle ahead, both circulating at the same speed. Infrared thermography indicated a fleeting contribution of traffic to RST. The temperature increase in circulated areas, with respect to uncirculated ones, does not last according to collected measurements. Measurements with atmospheric and radiometric probes provided elements to properly take into account traffic in a numerical model and to appreciate its contribution.