Problem: A common assumption is that students prefer to select their friends for group-work. The prime goal of this study was to understand the impact of two group selection methods on how students from diverse cultural backgrounds build learning and work relations. Method: Social Network Analysis in a pre-post test manner in a quasi-experimental design of 81 vs. 70 third-year students. Solution: In this study, we “disrupted” this group selection process after Day 1 by balancing students from different parts of the social network together. In one condition the students were “balanced” into groups by staff to encourage structural hole formation, and in the other condition students were allowed to self-select their group members to encourage network closure. Results: Students in the self-selected condition primarily selected their friends from a similar cultural background. In both conditions the learning networks after 11 weeks were primarily predicted by the group allocation and initial friendships. However, students in the balanced condition developed more cross-cultural learning links. These results indicate that teachers can actively intervene in the cross-cultural dynamics in- and outside the classroom.