Teams play a vital role in the workplace, enabling individuals to collaborate to solve complex and challenging issues (Salas, Shuffler, Thayer, Bedwell & Lazzarra, 2015). This is especially true for teams operating in extreme and challenging environments, in which decisions can have life or death consequences (Bell, Fisher, Brown & Mann, 2016). Our research is interested in teamwork during major emergency incidents, in which multiple component teams must work interdependently to deliver an effective coordinated response (Power, 2018). This is referred to as a multi-team system (MTS), wherein teams are working together to achieve separate but related objectives in the context of over-arching collective goals (Shuffler, Jiminez-Rodriguez & Kramer, 2015). Previous research has demonstrated the importance of team composition to support team performance. Specifically, evidence has demonstrated the benefit of familiarity amongst team members (Harrison, McGrath, Florey & Vanderstoep, 2003) and homogeneity of individual traits (e.g., personality and values) (Morgenson, Reider & Campion, 2005). However there has been little empirical evidence of how composition might support teamwork in MTS. Furthermore, researchers have theorised about how changing goal hierarchies and the needs of individual component teams can interact with MTS functioning (Shuffler et al., 2015), with little experimental evidence to test this. In this study we aim to explore how team composition and goal priorities impact the ability of an MTS to work effectively. Specifically, we hypothesise that familiar teams will communicate and coordinate more effectively during a simulated major incident and they will show more goal congruence in the latter stages of the simulation that non-familiar teams. In addition, we hypothesise that homogeneity in traits/values will also interact with team behaviour and that this may interact with familiarity.