Quantifying the effect of rail transit on vehicular traffic helps policy makers understand its transportation benefits. Previous studies seldom consider the effect over time and the influence of confounding factors. We apply a quasi-experiment research design to explore the evolving impact of the Green Line light rail transit on vehicular traffic in the Twin Cities, controlling for road classification, land use, and transit supply. The results show that rail transit is a substitute for automobile traffic, but induced and diverted trips gradually reduce the substitution effect. The reduced effect suggests that rail transit improves transportation system performance.