Electronic Toll Collection System (ETCS), so called "Hi-Pass" in Korea, has improved traffic flow at toll gate of highways. It is known that the improvement of traffic flow should reduce air pollutants and $CO_2$ from vehicles. In this study, real driving emission of a light duty truck with Portable Emission Measurement System(PEMS) has been measured to evaluate the emission reduction effect due to ETCS. The correlations between driving variables and emissions have been analyzed to verify its effect on traffic flow improvement and emission reduction at toll gate. We considered average vehicle speed, Relative Positive Acceleration (RPA), and the distance of queue as driving variables. Compared to passing Manual Toll Collection System (MTCS) lane without queue, ETCS was able to reduce 38.7% of $NO_x$, 21.6% of soot, and 27.7% of $CO_2$. The results showed that the higher the average vehicle speed, the lower RPA and no queue in ETCS contributed to the emission reductions. Linear equation models with RPA and queue have been established by the multiple linear regression method. The linear models resulted in the higher coefficient of determination than those with only average vehicle speed used for establishing vehicle emission factors.