As an effective travel demand management means, park-and-ride (P&R) mode is an important part of urban traffic. In a traffic corridor with P&R service, suppose that the travel time on highway is uncertain, a cumulative prospect theory (CPT)-based travel decision-making model is established with two travel modes of driving all the way and (P&R). With this setting, the effect of various factors such as the transit fare rate, the parking fee and the total travel demand on the CPT-based and expected utility theory (EUT)-based equilibrium results are compared. In addition, the sensitivity analysis focus on CPT-related parameters are also performed. The numerical results in our case show that the equilibrium flow on P&R mode is always underestimated in an EUT-based model, especially for a low total travel demand. Also, it is found that reducing the transit fare rate or parking fee for P&R station and raising the parking fee for CBD has the same effect on promoting more commuters transfer to P&R mode, whatever CPT-based or EUT-based model is employed.Furthermore, commuter’s reference dependence characteristic is also observed in a CPT-based model, and it is especially noticeable when the road uncertainty is large.