The complexity and uncertainty of the future, motivated by the evolution of technology in the transport system, opens space for different scenarios. One of the solutions, driven by the introduction of maximum vehicle automation in the market is the Peer-to-Peer Autonomous Car-Sharing (P2P ACS) service, i.e. a car-sharing service that uses autonomous vehicles from and to individuals. This research intends to assess the P2P ACS system, based on the explicit representation of demand and supply decisions in an agent-based framework (AnyLogic), and to quantify the potential advantages of P2P ACS in the operation of mobility and sustainable development in urban areas. Some pricing policies to avoid empty trips will also be considered in this study.