Connected and automated driving is expected to increase road safety. Safety impact assessment is an approach to prospectively assess the potential safety benefits of systems before their introduction to market. As these assessments usually leverage model-in-the-loop simulations, it is required to model the involved road users’ behavior realistically. Current research on prospective safety impact assessment mainly focusses on automated driving functions (ADFs) not taking vehicle communication into account. When assessing ADFs incorporating collective awareness of other road users based on vehicle communication, the effects of the communication are to be modeled in addition. Cooperative awareness aims at overcoming perception shortcomings of ADFs due to limited sensor ranges or visual obstructions, as road users communicate their own dynamic state. This work presents a methodology for prospective safety impact assessment of cooperative awareness-enabled ADFs. The methodology includes cooperative awareness data analyses, generating realistic baseline scenarios, as well as modelling the accuracies of the communicated information. The developed methodology is evaluated for vehicle-to-vehicle communication of ETSI Cooperative Awareness Messages (CAM) based on real-world data from the V2AIX dataset. This use case is assessed for different ADFs at urban intersections affected by visual obstructions. The results show an increase in the crash occurrence for realistic CAM modelling compared to ideal CAMs. This implies that assessing cooperative awareness in simulations requires modelling it carefully to achieve reliable results.