Recent advances in traffic flow theory at the network level, namely the Macroscopic Fundamental Diagram (MFD), reveals the existence of well-defined laws of congestion dynamics at aggregated levels. The same knowledge for multimodal networks however is limited. It is critical to understand how urban space can be allocated and managed for multimodality. The objective is to develop aggregated modeling and optimization approaches, which will contribute on the knowledge of congestion dynamics in cities of different structures and mode usages, and ultimately facilitate the design of efficient and equitable urban transport policies. Building on the knowledge of the single-mode MFD theory, a bi-modal MFD model considering the effect of mode conflict is proposed for mixed networks of buses and cars. A system-level model is developed for multiple-region city network. The flow dynamics among regions are described by a regional level flow conservation law. A non-linear optimization framework is performed to optimize space allocation, minimizing the total passenger cost, given certain demand, city structure and road facility. Then, parking limitation is integrated in the proposed multi-modal system model, where vehicles cruising for parking are also integrated. The extra delay of cruising is captured by a geometric distribution related to the time-dependent parking availability and estimated at the aggregated level. The delay cost to other users is also estimated via the bi-modal MFD, and it shows the effect of cruising on all travelers who do not require parking. Optimal parking pricing policies for on-street and garage parking are obtained through the optimization framework, as well. The existence of a three-dimensional MFD (3D-MFD) for mixed bi-modal networks is investigated and analyzed via micro-traffic simulation studies. A 3D-MFD relates vehicular production of a network (flow, travel distance) to the density of cars and buses, where the impact of each mode on network performance can be directly observed. To further compare the modal impact on performance, the Bus-Car Unit equivalent value is estimated, indicating that this value is state- and mode-composition dependent rather than deterministic. In addition to the conventional vehicle-flow-based analysis, a passenger 3D-MFD is derived which provides a different perspective of the flow characteristics in bi-modal networks. Simulation study on 3D-MFD based perimeter-control shows promising performance in real-time control. The final part of the thesis concerns the MFD-controlled congestion pricing. Feedback-type control mechanisms are proposed to determine and adjust the time-dependent tolls, based on congestion level as expressed by the MFD. One pricing scheme also considers user’s adaptation to the toll cost, allowing a great flexibility in toll adjustment, and deals with the promotion of public transport usage. The performance of the pricing schemes is investigated in an existing agent-based model where the complex travel behavior in real-life is reasonably reproduced. Results demonstrate that the pricing schemes are effective in congestion reduction. Remarkably, smooth behavioral equilibrium in long-term operation is found under such pricing schemes. Furthermore, user heterogeneity with respect to value-of-time is introduced in the agent-based model. By realizing and treating this heterogeneity, pricing strategies can achieve even higher efficiency and equitable benefit.