Replacing motor vehicle journeys with travel by foot and bicycle is recognised as a means to help achieve a range of health, environmental and economic objectives. Close passes - where a motor vehicle overtakes a cyclist with a minimal lateral clearance - have been identified as both a prominent safety concern and a barrier to increased uptake of cycling. Close passes are the most common type of on-road incident experienced by cyclists where a motor vehicle is involved and are also associated with collisions resulting in deaths and serious injuries. The current study builds on existing research investigating the underlying causes of close passes by examining responses from a sample of drivers to videos of close passes that were submitted by cyclists as driving complaints to an English county Police force. An online survey recorded the opinions of UK drivers (n = 293) on the behaviour of the cyclist and driver featured in each of 8 video clips. The survey recorded participants’ road use habits including the range of driving purposes engaged in (“driverXP”), their self-defined identity on a driver-cyclist spectrum, and their knowledge of recommended practice for bicycle road positioning. It also included the ‘Driving Anger eXpression’ (DAX) scale, which scores tendencies towards different forms of anger expression while driving, including use of vehicle manoeuvres to express anger (“vehicleDAX”). Multiple regressions were used to determine the relationship between these independent variables, and participants’ assessment of how risky the overtake was, and the extent to which the driver or cyclist were seen as liable for what happened.The level of liability apportioned to the cyclists featured in the clips was higher from drivers who were not cyclists themselves, compared with those who did cycle, and from drivers who scored higher on the vehicleDAX and driverXP measures (Identifies as cyclist τ = -0.377, p <.01), (driverXP τ = 0.209, p <.01), (vehicleDAX τ = 0.113, p <.05). Drivers with higher vehicleDAX scores showed lower levels of agreement with the statement “the incident was dangerous” in relation to the close pass clips. The liability apportioned to the cyclists featured was mediated by the drivers’ cycling positional knowledge score (ΔR2 = 0.139, p <.001).This is the first time that naturalistic footage reported by cyclists to Police has been used to examine differing perceptions of the same close pass event. Findings show a need to share knowledge about cycling practice in order to help drivers navigate cycle traffic safely and considerately. This can inform road safety interventions that can contribute to safer cycling and driver behaviours that are more welcoming to existing and potential cyclists.