Abstract: Although manufacturers and experts consider autonomous vehicles (AVs) as a much safer alternative than traditional human-driven vehicles, the lack of trust and high perceived risk by potential users can be a major obstacle to their acceptance. While both risk and trust have been the focus of interest for AV researchers, studies have often produced contradictory results. This study offers a new perspective to investigate the effect of trust and perceived risk to resolve these ambiguities. We identified three underlying dimensions of trust supplemented by two dimensions of risk and incorporated them into one model. The proposed model offers direct and indirect paths between trust dimensions and AV acceptance with the mediation of the dimensions of perceived risk. Based on a survey of 949 adult respondents, the model was tested with structural equation modeling (SEM). Results revealed that only performance trust affected directly intention to use AVs, while trust in manufacturers influenced intention to use with the mediation of privacy risk. An important result is that trust in institutions that can influence future rules and regulations for the use of AVs has no impact either directly or indirectly on intention to use. The practical implications can assist regulators and manufacturers to increase their efforts to build trust and confidence, thus enhancing the adoption of this technology.