Despite the prevalence of teams in research, there is a lack of a good understanding of how their size and diversity affects their performance. We develop a theoretical framework that distinguishes two dimensions of research performance for an academic paper: impact (i.e., subsequent citations) and prestige (i.e., ranking of the journal where research is published). We propose that, while larger teams will enhance linearly the impact of research, they will affect its prestige in a nonlinear fashion. We further contend that these effects will be moderated by knowledge and international diversity of the teams. We test these hypotheses using bibliometric data between 1990 and 2020 on more than 1.4 million papers and 18 million citation counts across 22 subfields in management. Our results confirm significant benefits for research impact from both team size and diversity, but also highlight drawbacks when teams become very large and heterogeneous. Moreover, we find a nonlinear positive effect of team size on research prestige that can be offset only by high levels of knowledge diversity. These findings are robust to a variety of proxies, controls, and estimation techniques, including instrumental variables and propensity score matching. We discuss practical implications for stimulating research performance in business schools.