Traditional models of transport networks ignore the existence and strategic behaviour of large actors who are often active in transport markets. Examples of such large actors are private infrastructure or service operators, insurance companies, or vehicle manufacturers. Both for positive and normative analyses, this omission can lead to substantial errors, and therefore to seriously biased policy evaluations and recommendations. The reason is that such actors will have their own objectives to pursue, while their market power gives them ample opportunity to influence market outcomes through strategic behaviour. Ignoring their behavioural responses to policy changes therefore leads to a wrong prediction of the policy s optimal design as well as its impacts. An important reason why they are nevertheless usually ignored is the analytical and numerical complexity of transport network models in which large actors, with strategic behaviour, are active. This project seeks to develop such models. Specific applications will include models of road transport networks allowing for private road operators, vehicle manufacturers, and insurance companies; models of urban taxi markets; and networks models for public transport and aviation. Although applying to different cases, these models will have important methodological characteristics in common, particularly in that they apply multilevel optimization techniques for (transport) network models that account for strategic behaviour of and interactions between large actors. We will investigate how this behaviour affects the formation of network equilibria in transport markets, as well as the impacts and (second-best) optimal design of transport policies.