When assessing learning outcomes of role-play simulations of political decision-making, research points to inconsistencies and shortcomings. Different learning environment components aren’t often taken into account and little attention is given to affective learning outcomes and the development of student learning during the simulation. Therefore, this study focuses on the development of self-efficacy for negotiating during a simulation and includes both student characteristics and simulation features to explain variation. Data were collected during a four-day simulation of decision-making and self-efficacy for negotiating was measured at four time points. Gender, simulation experience and opportunities to practice negotiating during the simulation appear to play a role when assessing how students vary in their self-efficacy. However, findings revealed no statistically significant on average development of self-efficacy. Simulation experience and opportunities to practice negotiating, and its impact on students’ learning processes and outcomes should not be underestimated and deserves to be investigated in more depth.