Objectives: Physicians with different professional backgrounds often diagnose a patients’ problem collaboratively. In this article, we first introduce a process model for collaborative diagnosing (CDR model), describe the development of a simulation used to empirically examine the facilitation of collaborative diagnostic reasoning. Based on a contemporary validity framework [1], we further suggest indicators for validity and collect initial evidence with respect to the scoring, generalization, extrapolation, and implication inferences to assess the validity of the simulation when used to assess effects of learning interventions. Method: In a quasi-experimental study, we assessed objectivity and reliability of the simulation and compared medical students with low and advanced prior knowledge to practitioners with high prior knowledge with respect to their diagnostic accuracy, diagnostic efficiency, information sharing skills, and their intrinsic cognitive load. Additionally, we obtained authenticity ratings from practitioners with high prior knowledge. Results: The results yielded satisfying initial evidence for the validity of the scoring and the extrapolation inferences as ratings are objective, and the simulation and the collaborative process is perceived as rather authentic. Additionally, participants on different levels of prior knowledge differ with respect to their diagnostic accuracy, diagnostic efficiency, information sharing skills, and their reported intrinsic cognitive load. With one exception (information sharing skills), the generalization inference seems to be valid as well. Conclusions: We conclude that collecting validity evidence for the simulation was an important step towards a better interpretation of the simulation. We found that the simulation is an authentic and valid representation of the chosen collaborative situation and that the collected validity evidence offers sufficient evidence for an initial validation of the simulation. Nevertheless, the validation process highlighted some important gaps that need further consideration. We further conclude that applying a validation model to the context of empirical research is promising and encourage other researchers to follow the example.