A novel, discrete space-time model of pedestrian behaviour in real urban networks is presented. An agent-based approach is used to define characteristics of individual pedestrians, based on spatial awareness and cognition theories, combined with preferential choices of different social groups. Behaviour patterns are considered incorporating rules of movement along pedestrian routes and for intermediate decision and conflict points. The model utilises dynamic volunteered geographic information system data allowing analysis of arbitrary city networks and comparison of the effect of grid structure and amenity distribution. As an example, two distinctive social groups are considered, namely 'directed' and 'leisure', and their interaction, together with the way in which flow congestion and changes in network morphology affect route choice in central London areas. The resulting stress and flow characteristics of the urban network simulations as well as the impact on individual agent paths and travel times, are discussed. © 2014 Springer International Publishing.