Background Against the backdrop of the cognitive-motivational process model proposed by Vollmeyer and Rheinberg (1998) , this study investigates how the personality trait need for cognition in combination with a specific situation (short and long version of a live-action simulation game), current motivation, and immersion and flow (mediators) predict the learning outcome in the simulation game EVERYDAY LIFE IN THE CLASSROOM ( Starker & Imhof, 2014 ). Methods N = 138 teacher students played the short 2-hour version of the simulation game, and N = 77 played the long 2-day version. Need for cognition, current motivation, immersion, flow, and learning outcome were measured by self-report questionnaires. Results A hierarchical regression analysis was used to determine the impact of the predictor variables (need for cognition, current motivation, immersion, and flow) on the self-evaluated learning outcome. For the 2-hour group, the predictor variables interest, challenge, and immersion (subscale: engagement) impacted self-reported learning outcome significantly positive. For the 2-day group, significant predictors were need for cognition, interest, and immersion (subscale: engrossment). In comparison of the two game versions, students playing the 2-day version reported significantly higher levels of engagement, engrossment, and learning outcome. Conclusion The cognitive-motivational process model of learning was partly supported: interest and immersion predict learning outcome in the live-action simulation game. The extended 2-day version of the game leads to higher levels of immersion and higher learning outcome, indicating that a longer timeframe secures the desired effects on learning outcome from simulation games. Further research needs to shed light on the interaction of personality traits and immersion.