Average-ability students often do not take full advantage of learning in mixed-ability groups because they hardly engage in the group interaction. This study examined whether structuring collaboration by group roles and ground rules for helping behavior might help overcome this participatory inequality. In a plant biology course, heterogeneously grouped fourth-grade boys (n=164) were randomly assigned to a structured collaboration condition or an unconstrained comparison condition. Results indicated positive effects of structured collaboration on average-ability students' achievement, motivation, and contribution to the group interaction. Another positive result was that structuring collaboration did not lower the scores of the high and low-ability students on these measures.