Purpose – The purpose of this paper is to introduce and test a model of group processes (e.g. conflict), emergent states (e.g. trust), and group context (e.g. connectedness) to better understand the mechanisms that underlie the traditionally negative effects of conflict. Design/methodology/approach – A total of 27 workgroups of a Dutch telecommunications company participated in a survey. To assess trust as a mediator between conflict and performance bootstrapping analysis was used. In addition, the moderating role of the three connectedness types was investigated with hierarchical regressions. Findings – The results suggest that trust partially mediates the effect of task conflict and fully mediates the effect of relationship conflict on performance. Furthermore, trust is less affected by task conflict when group members are highly cognitively connected and less affected by relationship conflict when group members are highly task connected. Research limitations/implications – This research implies that task and cognitive connectedness decrease the negative effect of conflict on trust, and hence, performance. Shortcomings include discussing the causality between conflict and trust, and the possibility of different perceptions among group members regarding group phenomena. Practical implications – The findings suggest that managers can help to provide circumstances in which conflict is not necessarily destructive for intragroup trust and performance. Originality/value – Provides one of the first empirical examinations of the mediating role of trust in the relationship between task and relationship conflict and perceived group performance. Additionally, examines if connectedness (the level of active involvement of group members with each other) buffers the negative effects of conflict on trust.