This study systematically integrated empirical literature on the relationship between team trust and team performance focusing on real educational and organizational teams. A total of 84 empirical studies comprising 84 independent effect sizes involving a total of 5,582 teams met the inclusion criteria. Applying a random-effects model, the overall effect size between team trust and team performance was positive and moderate (mean r = 0.34). After correcting for measurement unreliability, the overall effect size increased (mean r corrected = 0.40). The between-study heterogeneity was significant (Q = 385.30 [95% CI, p < 0.00), suggesting the existence of moderating variables. Nine potential moderating variables were examined including: team size (large versus small), team tenure (ad hoc versus intact), mode of communication (face-to-face versus virtual), task complexity (high versus low), study team setting (educational versus organizational), study age (early versus recent), cultural context of the sample (collectivistic versus individualistic), type of team performance measure (objective versus subjective) and operationalization of team trust (aggregation versus key informant). The results of these moderators are discussed along with the potential for publication bias. Limitations and suggestions for future research are also discussed.