The purpose of this article is to explore the relationship between team diversity, conflict, and performance with the meta-analytic method. This paper quantitatively describes the overall condition of relevant research on mechanism among team diversity, intragroup conflict and team performance in recent years, analyzes moderators between intragroup conflict and team performance and establishes a contingency model of team diversity-conflict-performance so that a clear mechanism is drawn. Nowadays, it is vital for organizations to understand the mechanism of team diversity and make full use of it. However, the relationship between team diversity and team performance is not consistent in relevant research. In recent years, several meta-analytic results indicate that team diversity is a potential predictor of intragroup conflict and conflict can predict team outcome such as team performance, team member satisfaction and cooperation to a certain degree. However, relevant meta-analyses mainly focus on surface–level team diversity such as informational diversity and background diversity and rarely consider deep-level diversity. Furthermore, to the best of our knowledge, the relationship between team diversity, conflict and performance is rarely explored with meta-analytic method. In addition, several important moderating variables and contextual factors need further analysis. Under such circumstance, based on input-process-outcome model, the present paper analyzed 112 empirical articles (n=9263 groups) both home and abroad using the meta-analytic method. The meta-analysis was used to aggregate results from studies examining the relationship between team diversity, conflict and team performance and also investigate a number of moderating variables. We searched thoroughly using relevant key words such as team diversity in combination with conflict in databases and engines both home and abroad(Google Scholar, EBSCO, ABI/Inform, Proquest, Elsevier, CNKI, etc.), top journals in this field and even unpublished articles and working paper between 1990 and 2012. After screening, 112 empirical articles based on the responses of 9236 groups met the inclusion criteria in the meta-analysis. We coded the studies and analyzed the data collected with the Comprehensive Meta-Analysis( CMA) Version 2 program. Results showed that: a) team deep-level diversity was an important factor leading to conflict in team, but surface diversity wouldn’t increase conflict much; b) unlike research in USA, research in China showed that task conflict reduced team performance significantly, and task conflict didn’t differentiate much from relationship conflict; c) subject type, response rate, team size and the asymmetric perception of conflict were five moderators moderating the relationship between team conflict and performance. Several conclusions are drawn: team diversity is an important predictor of intragroup conflict; contextual factors such as culture and asymmetric perception of conflict are able to moderate the relationship between conflict and team performance. We also describe the theoretical and practical implications of these findings for advancing relevant study.