Road authorities struggle with the question whether variable message signs (VMSs) should exclusively be used for traffic management or could also be used to display traffic-irrelevant messages, such as mottos or commercial advertisements. The current study assesses behavioural responses to a critical route instruction displayed on the same VMS that previously displayed a variety of traffic-irrelevant messages. For this, thirty-two participants were divided between a control group and an experimental group (the advertisements group). In a driving simulator, all were familiarised with the same route by driving a VMS-equipped motorway nine times. For the advertisements group, up to drive 8, this VMS displayed various advertisements. Whereas for the control group it was blank. In the 9 th drive, the VMS displayed a critical detour message for all participants. This critical route instructioninforming drivers to take the nearest exitresulted in compliant driver behaviour in the advertisements group. In addition, they only reduced speed marginally to increase the time to process the VMS text. The control group, on the contrary, displayed a much sharper speed reduction; though the instruction only moderately altered motorway exit behaviour. What is more, the 31% (n = 4) of the advertisements group who complied with the critical route instruction subsequently failed to recall this message (recalling an advertisement instead). In conclusion, this study provides evidence that displaying traffic-irrelevant messages on VMSs might not interfere with traffic management; provided the format of said messages is in accordance with ergonomic VMS guidelines as used in this study. It is proposed that due to repeated exposure to various VMS texts, reading the sign has been practised to the extent that little to no conscious deliberation was required. As a result, recall of what was seen, proved to be an inadequate proxy for assessing driver behaviour. This study shows that conscious attention might not be a prerequisite for compliance. Furthermore, it suggests that continuous variability in objects in the traffic environment may become part of a subconscious monitoring process, as long as they have been sufficiently practised.