Guangzhou has over 12.7 million inhabitants, is located in the south of mainland China and is the capital of Guangdong Province. By the end of 2010, there were 2.14 million motor vehicles in exceeds all other complaints. Therefore, day and night road traffic noise mapping in Guangzhou should be developed. However, there are two critical challenges. First, it is extremely difficult to obtain basic traffic information (e.g., the traffic volume, the vehicle speed, etc.) and configurations (the attributes of roads and buildings) for a large city. Second, the regional traffic noise calculation algorithm has a very low efficiency. In this study, we will present several ways to address these problems. For example, GPS data collected from floating cars are used to estimate the speed and traffic volume of the road network; the attributes of roads and buildings are exported from GIS automatically; and the algorithm is optimized to effectively evaluate road traffic noise levels. In addition, the results are projected onto a GIS map to create day and night traffic noise maps of Guangzhou, and the accuracy of the algorithm is verified using a traffic noise monitoring experiment.