How do active learning environments – by means of simulations – enhance political science student’s learning outcomes? And what are the differential effects according to specific student attributes such as gender or prior knowledge? This paper examines different UN simulations in political science courses in order to provide proof of this learning tool’s effectiveness. In contrast to theoretical claims about the positive effects of active learning environments on student learning outcomes, substantial empirical evidence confirming these claims is still limited. Here, we focus on one specific tool, simulations, in order to systematically test previous claims and demonstrate the pedagogical value of using simulations in political science curricula. Model United Nations (MUNs) are a type of simulation that has been a popular device in teaching political science. To gain comprehensive data about active learning effect of MUNs, we evaluate simulations at three different levels of complexity. First, on the lowest level of complexity, we assess a simulation of the UN Security Council, which runs for three hours and is an in-class simulation. Second, we survey MainMUN, a four-day simulation taking place in Frankfurt, Germany. With different committees being simulated and an individual preparation time, it has a medium level of complexity. Finally, on the highest level of complexity, we evaluate delegations of the National Model United Nations that takes place in New York. In our case, students prepare for one year to take part in this five day simulation. At this highest level of complexity, we conduct a longitudinal study, using data from two consecutive years. Preliminary results prove that simulations as a teaching tool in political science have positive effects on the learning-outcomes of the students in different ways: e.g., they lead to enhanced knowledge about the UN and boost soft skills and the ability to reflect on different positions.