<p dir="ltr">Breakdowns in communication and teamwork are significant contributors to errors in safety-critical domains such as healthcare and aviation. Virtual Reality (VR) simulations have emerged as promising tools for team training in these environments, offering immersive experiences at lower costs compared to traditional high-fidelity simulators. However, current team evaluation methods often rely on subjective assessments and individual-level metrics, which may not fully capture the complex dynamics of team interactions. This thesis addresses the need for objective, team-level metrics in VR-based simulations by developing a collaborative three-person VR task designed to elicit varying levels of team workload through manipulations of task interdependence and time pressure. Data were collected from 60 participants (20 teams) using subjective questionnaires, objective performance metrics, and team-level physiological measures. The findings demonstrate that the experimental tasks effectively elicited distinguishable differences in team performance and perceived workload across varying conditions. Subjective measures of team workload correlated moderately with objective performance metrics, supporting the validity of the performance measures. While MdRQA metrics did not significantly predict subjective workload perceptions, they were able to predict performance metrics including error rates, suggesting that team-level physiological metrics are associated with team performance and may be predictive of strategy shifts in VR tasks. These results reinforce the feasibility of using VR simulations for evaluating team performance and highlight the potential of integrating objective physiological metrics into VR training environments. This integration can strengthen the assessment and improvement of team performance in safety-critical domains, offering a more comprehensive approach to evaluating teams.</p>