This article analyzes the relationships between highway capacity additions and transit patronage, both in the short and long run. A methodology using a model of schedule disutility is shown to provide a technique to account for transit service frequency. This technique, combined with a supply-side model of a highway corridor is used to evaluate the impact of transit headway changes and highway capacity, increases on total transit ridership, using a synthetic sample of commuters. Simulation results are used to evaluate the impact on travel times and utility of the two modes and the long run degradation of transit service predicted by the Downs-Thomson paradox. While the results do not show congestion as necessarily being worse than before capacity expansion, they do show that transit service frequency could be reduced significantly over time.