We propose a new link-based, bottom-up vehicle emission inventory method for Chinese cities using available on-road emission measurement data and activity survey data. This study uses Beijing as a case study and focuses on CO, HC, and NOx emissions from hot-stabilized activities of light-duty vehicles. The target year is 2004. The total quantity of emissions, emission intensity, temporal distribution of emissions by hour, and spatial distribution of emissions at 1 km x 1 km resolution are presented and compared with results from other inventory methods commonly used in China. The results show that the total emissions from light-duty vehicles in the Beijing urban area in 2004 were 1141 Mg of CO per day, 48 Mg of HC per day, and 32 Mg of NOx per day. Pollutant emissions from mobile sources show temporal and spatial variation trends that are consistent with the characteristics of people's activities. The inventory developed in this study reflects the influences of vehicle driving patterns in traffic on vehicle emissions on a road link basis, and it could be used to evaluate the impacts of traffic-related emission control measures in China.