The development of automated driving has made huge advances in the last years. Several companies have announced the deployment of highly automated vehicles in public traffic in the near future. However, the operation of SAE level 3+ automated vehicles is inherently subject to risk caused by uncertainty in environment perception, prediction of other traffic participants’ behavior, vehicle behavior generation, requirements specification, and validation as well as managing a corresponding very complex computer system. Hence, novel approaches towards ensuring and validating safety of automated vehicles are necessary in order to reduce the risk to an acceptable level, which is itself not defined or even quantifiable yet. Thus, the third workshop on “Ensuring and Validating Safety of Automated Vehicles” addresses multiple topics related to the safe operation of automated vehicles. Among others, topics such as safety concepts, hazard analyses, handling of challenging and critical scenarios, impact of machine learning on safety argumentation, as well as concepts for validating the functionality of automated vehicles are welcome contributions.