This dissertation explores and analyses the challenges and needs that developments in the information society are bringing to knowledge production supporting policy development and strategic decision making in the field of transport. Currently, the context of transport policies is about to shift from a transport infrastructure network design towards the development of a large socio-technical system, depending largely on ICT technology and applications. Dynamic decision making clusters or networks, consisting of different actors and having a variety of goals, are growing around policy items or transport system innovations, and they need information and knowledge as the basis for their mutual decisions. This development will change the roles of the different actors within the system as well as the nature of strategies and measures. My key argument is that in this new context the traditional, analytical knowledge production approaches (such as "planning" and "impact assessment", referring to infrastructure investments and project appraisals) are alone not sufficient in providing the knowledge needed to understand the socio-technical nature of the transport system or the dynamics between the different actors, as a basis for transport policy development. The knowledge provided to make informed transport decisions needs to include, in addition to the traditional issues, also new forms to serve the needs of a wider variety of societal actors. Based on the field of science and technology studies (STS), which aims to illuminate the relationship between knowledge and political power as well as investigating the place of science and technology in society, I have identified five emerging forms that I consider relevant to transport policy knowledge production in the future. These are knowledge production through system-based foresight, knowledge production through system-based evaluation, knowledge production in networks, knowledge production as processes of social learning and