Abstract:Unlike previous studies, the stability of research teams is measured from team level, while focusing on the moderating role of disciplinary diversity in this process. Using 69 biology academician teams of United States National Academy of Sciences in 2009 and 2010, the relationship between team stability and team performance and the moderating role of disciplinary diversity are analyzed using negative binomial regression models, based on the identification of key members and teams through author co-occurrence networks and cluster analysis, and team performance measured by team productivity and team influence. It is found that team stability has an inverted "U" shape effect on team performance, indicating that a more stable team does not equal to a better team, and that appropriate team member turnover is more conducive to team performance. Disciplinary diversity has a positive moderating effect on team stability and productivity relationships because it reduces the learning costs and time costs associated with team member mobility, making the impact of member mobility on team performance more significant; while disciplinary diversity has a less significant moderating effect on team stability and influence relationships because it tends to cause team knowledge conflicts and fragmentation of team resource allocation. To improve the performance of research teams, it is important to pay attention to the mobility of team members, optimize team entry and exit mechanisms, and develop more interdisciplinary talent collaboration.