Transporting parcels on urban passenger rail transit is gaining growing interest as a response to the increasing demand and cost of urban parcel delivery. To analyze the welfare effects of different fare regimes when allowing parcel services on an urban rail transit, this paper models the optimal service problem where the transit operator chooses the number of trains and the departure intervals. By introducing a reduced form train timetable problem, the passenger train crowding model is extended to incorporate the effect of freight train scheduling. We show that the freight users are better off in the time-varying optimal fare regime, while passengers are worse off, and that the time-varying optimal fare regime calls for more trains than the optimal uniform fare regime. However, the reduction in passenger trains due to the introduction of freight service can eliminate the welfare gain from passenger time-varying fare. If the price elasticity of freight demand is relatively high, implementing road toll can generate welfare loss when rail transit is privately operated.