AimThe increasing use of high-fidelity simulation is limited by the imbalance between the growing number of students and the human resources available in such a way that all residents cannot play a role during scenarios. The learning outcomes of observers need to be studied. MethodsThis prospective randomised study was approved by the institutional review board. Anaesthesia residents attending a one-day training session were enrolled. In each of the four scenarios, three residents played an active role while others observed in a separate room. All participants attended debriefing sessions. Residents were randomised between active participant-observer group (AP-O group) and observer group (O group). A similar questionnaire was distributed before, immediately after the session and after three months and included self-reported assessment of satisfaction, medical knowledge (noted 0–16), and non-technical skills. ResultsA hundred and four questionnaires were analysed. Immediately after the simulation, a significant increase in medical knowledge was recorded but was higher in the AP-O group (6 [5–8] to 10 [8–11]/16) than in the O group (7 [5–8] to 9 [7–10]/16). High scores for non-technical skills were similarly observed in both groups. Satisfaction was high in both groups but was higher in the AP-O group (9 [8–9] versus 8 [8–9]/10, P=0.019). Decay of knowledge was observed for most main outcomes at three months. ConclusionThis study suggests an immediate improvement of learning outcomes for both roles after immersive simulation but some learning outcomes may be better for residents engaged as players in scenarios.