The paper reports two experiments conducted within the Driver Behaviour and Traffic Safety research strand of the European DRIVE programme. Each experiment was conducted using members of the general public, driving specially adapted road vehicles, in real urban environments. The first experiment, conducted in Loughborough can be seen as a validation of a multi-level evaluation methodology developed within the project as applied to two modes of route information presentation: paper map or text display on LCD screen. The second experiment used similar data collection techniques, but was conducted in Berlin with two real route information systems (LISB and Bosch Travelpilot). Assessing the results of these two experiments shows that there is a greater degree of visual workload associated with drawing information off a map display. This is the result of analysing a combination of measures, including vehicle control data, physiological stress indices, visual attention data and a variety of subjective response measures. This form of information presentation may be less than compatible with the task of navigating a vehicle whilst interacting with the road infrastructures of today.