In normal, everyday driving, behaving safely in traffic usually remains unrewarded whereas consequences for unsafe behaviours are rarely negative and can actually often be positive. The strength of Pay-As-You-Drive insurance is its potential to correct this major issue with driving by rewarding safe and eco-friendly driving behaviour. It is hypothesised that the potential to provide useful feedback is strongest for Pay-As-You-Drive insurance systems that make use of in-car feedback systems. In-car feedback increases the certainty and swiftness with which target behaviour can be detected which is in line with good behavioural guidelines for providing feedback and rewards. The aim of the study presented here was therefore to compare a Pay-As-You-Drive system that provided immediate feedback while driving in a driving simulator, with delayed feedback provided through a website. For this purpose, participants (60 in total) were invited for two driving sessions and requested to drive for about 50 minutes in each session. The in-car and web group were told that they would receive 15 euro plus six euro depending on how they drove. Results showed that driving behaviour did benefit from both the in-car and web feedback linked to a monetary reward system, as the participants (twenty in each group) displayed less harsh cornering, acceleration, and speeding behaviour compared to the control group. As expected, immediate feedback resulted in a faster change in driving behaviour, although this initial difference largely disappeared after the web-group had received feedback. Nonetheless, it is concluded that the in-car system benefitted driving behaviour the most, since it was observed that vehicle parameters that were not directly involved in establishing the monetary savings, such as accepted gap time and time headways also improved. This indicates that for specific situations, direct feedback results in safer driving.