This study investigated the effect of different gender and ability composition on students' behaviors, interactions, and learning outcomes during small cooperative group activities. The study involved Year 6 children (N=440) who were assigned to one of 12 four-person groups that were either gender-balanced or gender-imbalanced with different compositions of high, medium, and low ability students. The children worked in their groups on a set of social studies activities for one hour each time three times per week for six weeks. Each group was videotaped twice during the period, once in Week 3 and again in Week 6. The results showed, contrary to expectations, that the effect of different ability and gender compositions was minimal. As the members of each group had more time to work together they became more responsive to each other's needs and provided more help and assistance to each other so that all groups attained comparable learning gains. (Contains 23 references.) (Author/ND)