Accidents involving vulnerable road users (VRUs), such as cyclists, remain a major issue for road safety, accounting for almost 40% of road fatalities in Europe, and almost 50% worldwide [1]. Active safety systems have the potential to mitigate the risk to cyclists by warning drivers and providing braking and/or steering interventions if a cyclist is present within their vehicle's path, and/or a collision is likely based on the current speed and trajectory of each party. How-ever, current systems can be prone to issuing high numbers of false declarations, due the in-herent difficulty in predicting intent; this can annoy drivers, encouraging them to ignore the system, find creative workarounds or deactivate it completely [2]. The PROSPECT project [3] aims to significantly improve the effectiveness of current VRU safety systems (in particular, for cyclists) by expanding the scope of scenarios addressed, and improving overall system perfor-mance through enhanced VRU sensing and advanced system control strategies. Nevertheless, false activations remain inevitable due to the dynamic behaviour of both parties. The nature and frequency of false activations and drivers' understanding of the system's intentions are therefore important, and likely to play a significant role in shaping drivers' trust and ac-ceptance, and their ultimate uptake of the technology. Acceptance testing is often limited to surveys or short-term, single-exposures to 'the system'. Strictly speaking, such studies measure acceptability (the driver's intention to use the system) rather than acceptance (how the driver incorporates the system into their driving, based on their experience of actually using it); this is a particular concern given that high acceptability does not necessarily predict high acceptance. In contrast, in a longitudinal simulator study, fifty participants undertook a 20-minute urban journey, on each of five consecutive week-days (representing their daily commute). On day five, the vehicle identified an approaching cyclist (obscured to the driver), and predicted their intention to enter the roadway into which the car was driving. As a consequence, the system activated an emergency braking manoeuvre, shortly preceded by an audible alarm (emulating the functionality of the new PROSPECT system). In a balanced between-subjects design, the behaviour of the cyclist varied: they either stopped at the road-side or continued to cross the road, giving rise to a false-positive and true-positive alarm, respectively. Driver acceptance was subsequently assessed at the end of the week based on the testing protocol developed as part of the PROSPECT project. Acceptance ratings, driving performance, and the driver's inclination to use or purchase the system following false activation, will be reported. Findings will subse-quently be used to understand the likely acceptance and impact of the PROSPECT active safety system on driver behaviour and cyclist safety.