“The goal of the European Transport Policy is to establish a sustainable transport system that meets society’s economic, social and environmental needs…” (ECE, 2009). This statement indicates the challenges that the European transport policy makers are faced with when facilitating an increasing freight transport demand with limited transport infrastructures. The development of an interconnected intermodal transport system has been recognized by the European Commission as an important, strategic task that will contribute to solving the dilemma between the accommodation of an increased freight flow and the need for a sustainable living environment. This thesis focuses on model-based, quantitative analysis for infrastructure network design decisions for large scale intermodal transport systems.. The involvement of public concerns, as represented by the governmental objectives on sustainability, brings additional complexity into infrastructure network design. Governments are often concerned with network design on a regional scale or a national scale. The enlargement of the network scale to an international level further increases the level of heterogeneity of the network, among other factors in terms of the number of actors involved, the diversity of transport demand and the variety of transport service supply. These new objectives and dimensions pose new challenges to freight transport infrastructure network design. This thesis proposes a new model to support policy making for an intermodal freight transport network. The model is able to simultaneously incorporate large scale, multimodal, multi-commodity and multi-actor perspectives. It can be used for integrated policy, infrastructure and service design. Results can be visualized per transport mode and per commodity value group on a geographic information system at segmental level, terminal level, corridor level, regional level, national level, and network level. Implementation of the model for a realistic scale network design is another contribution of this thesis. To this end, we calibrated the model by using two approaches: a Genetic Algorithm based method and a feedback-based method. The model was validated by comparing the modelled link flows with observations, testing the cross elasticities of the costs to demand and comparing the catchment area of the terminals with areas observed in practice. The calibration results indicate that the model adequately captures the network usage decisions on an aggregated level. The model was applied to Dutch container transport network design problems. Databases of Dutch container transport demand, features of the European multimodal freight transport infrastructure network, information about selected inland waterway transport services, and information about transport and transhipment costs, emissions and external costs were embedded in the model. After completing the theoretical and empirical specification the model was applied to policy decisions on the Dutch container transport. The thesis extensively discusses the integrated infrastructure, service, and policy design that may contribute to managing the costs of the freight flows, meanwhile ensuring a sustainable living environment. The main findings from the application are as follows. - A higher CO2 price can results in lower total transport costs, despite extra handling costs in intermodal transhipments. The costs saved by bundling freight and using intermodal transport can compensate the additional handling costs. As these cannot compensate for the internalized CO2 emission costs, the total operational costs borne by transport operators will increase. - Network efficiency can be increased by closing terminals that are not able to attract sufficient volumes of demand. However, it is not likely to happen in practice, due to the fact that the private terminal operators and the local governments have local interests to protect on those small terminals that may conflict with the objective of minimizing total network costs. - The hub-network-services assumed and tested in this study cannot compete with road transport or shuttle barge transport services in the base scenario due to the extra transhipment costs, low load factor, and low demand for IWW container transport. In a future scenario, these services are only feasible under very high traffic growth. - There is not one single optimal future infrastructure network. Instead, a good infrastructure network design mainly depends on the future demand, transport price, and development of new transport technology. Based on the conclusions drawn in this thesis, implementing the combination of CO2 pricing and terminal network configuration is more effective than solely implementing CO2 pricing, with regard to total network CO2 emissions. A range of efficient networks, forming a frontier of minimal total network costs and total network CO2 emissions, is presented in the thesis, instead of one single optimal solution. The frontier provides more options in terminal network optimization in terms of the target network performance. The question which is the optimal network will depend on the relative value placed on CO2 emissions. The thesis ends with a vision on future freight transport network design models. A potential research direction is to incorporate the dimension of time into the model. This extension will enable the model to capture dynamic demand; to be applicable for scheduling synchronized intermodal transport services; to provide more realistic estimations of transport emissions and to analyse network reliability, including network robustness and service robustness. Reference: CEC (2009) 'COMMUNICATION FROM THE COMMISSION: A sustainable future for transport: Towards an integrated, technology-led and user friendly system', Commission of the European Communities, Brussels.