Congestion on urban main roads occurs due to a limited infrastructural capacity as compared to the supply of vehicles. Two action plans exist to reduce the magnitude of congestion issues on urban main roads: reducing the supply of vehicles and thus the demand for infrastructure, or expanding the size of the network’s capacity by constructing new roads. Since space scarcity is opposing a problem in densely populated urban centers, the first option is considered a more promising alternative in this context. Reducing the supply of vehicles can be achieved by combining trips and increasing the average vehicle occupancy. Factors and policies are studied that affect the commuter’s travel mode. Incentives for carpooling (sharing vehicle capacity) and the link with travel time (uncertainty) savings caused by a reserved high-occupancy-vehicle lane are researched. Travel time, waiting time and trip costs have the most (negative) impact on the probability that carpooling is chosen as the travel mode. A tool is created that is able to predict the proportion of commuters that will carpool when certain physical facilities, policies or a combination of the two are in place.