Introduction The degree of emotional activation required for optimal learning in either hands-on or observer roles is unclear, as is the level of stress that impedes learning. Measuring emotional activation is time-consuming, and many scales measure threat or anxiety without considering pleasurable activation. This study examined emotional activation in the observer and hands-on roles in 2 different scenario designs. Methods This study was a 2-cohort, parallel study of graduate nurses and doctors completing 2 different courses in managing the deteriorating patient. We examined emotional activation by role across 2 scenario designs. We measured emotional activation on 3 anchored measures scales: the State Trait Anxiety Inventory, Cognitive Appraisal Index, and the Affect Grid with data analysis using analysis of variance and repeated measures. Results Hands-on learners experienced higher anxiety, threat, and arousal levels and less pleasure than observers in both scenario designs. There were no differences in pre-emotional and postemotional activation in immersive scenarios for either role and increased arousal and decreased threat and anxiety in the hands-on role in the pause-and-discuss scenario design. Conclusions Hands-on learners were more emotionally activated than observers in both scenario designs. There was significant perceived anxiety, threat, and pleasurable arousal in both roles and both scenario designs. Pause-and-discuss scenarios demonstrated similar levels of activation as the immersive scenario design. The Affect Grid provided a quick subjective view of arousal and pleasure in simulation participants, potentially providing educators with an indication of whether emotional activation is positive (excitement) or negative (stressful) and may be helpful in educational planning and future research.