In a driving simulator experiment, effects were tested of an integrated support system on driving behaviour, user acceptance and workload. An alternative workload measure was used based upon peripheral vision. Two modes of support (tactile and speech messages) were compared to a control condition without support. Subjects were confronted with critical incidents dealing with lateral and longitudinal control on rural roads and motorways. The experiment showed that speech warnings are better suited for things related to law-enforcement, while tactile warnings are more suitable for situations directly related to driver safety. Although tactile warnings do result in compliance in enforcement type of situations, compliance is less than with speech warnings. For safety-related situations that do not require an immediate response, such as when approaching a sharp curve while speed is too high, speech warnings have the same effects as tactile warnings. However, if the criticality of the situation increases fast, for example when a lead vehicle brakes unexpectedly, tactile warnings result in more favourable effects than speech warnings. Still, both speech and tactile warnings do have a favourable effect of driver safety in a general way. The Peripheral Detection Task showed to be sensitive to variations in workload that were caused by different driving situations and to variations caused by short lasting increases in workload as a result of attending to the system messages.