The goals of this research project were to develop a game simulation model which supports learning knowledge management in a game environment and to investigate the validity of that model. The validity of the model is approached from two perspectives: educational validity and representational validity, and, in addition, possible relationships between those two types of validity. The representational validity of the model is investigated in an explorative study with subject matter experts in the knowledge management field. The outcomes of the study reveal that the model possesses the representational validity to a large extent. In two experimental studies with university students, the influence of two factors, fidelity level of the model and mode of playing, on the educational validity of the model and perceived educational validity of the model is investigated. The results suggest that the model which possesses a higher level of functional fidelity and includes variables describing knowledge processes provides a higher level of educational validity, while the model with a higher level of physical fidelity or the model which excludes variables describing knowledge processes, provides a higher level of perceived representational validity. A collaborative mode of playing, when compared with individual playing, provides a higher level of educational validity and no effects of collaboration on perceived representational validity are found.