With increasing urbanization and a need to reduce greenhouse gas emissions, new forms of urban mobility are studied worldwide. An interesting transportation method is ridesharing, where people can share parts of their rides with other passengers travelling in a similar direction in the same car. To im- prove the attraction of ridesharing systems, research into extensions to the existing ridesharing model formulation is required. The goal of this research is to make ridesharing more attractive to potential users by including the option to make reservations. In this work, the inclusion of reservations (pre-bookings) into the on-demand ridesharing problem is studied. The inclusion of ridesharing has to potential to make better assignments by having knowledge on part of the future demand. The main challenge from reservations is the increased computational load generated by having more available information. This thesis proposes multiple adjustments and different heuristics for a state-of-the-art on-demand ridesharing system to deal with the requirements and challenges brought by the incorporation of reservations. An additional consideration is a trade-off between providing benefits for reservations and serving on-demand requests. Numeric experiments are performed on the Manhattan street network with NYC taxicab request data. The results indicate that reservations can be included in on-demand ridesharing at a tolerable compu- tational cost with the proposed framework. Passengers that make reservations shortly in advance can be guaranteed service once initially accepted by the system. Having passengers make reservations in advance does increase the overall service the system can provide.