Inter-class ability grouping, i.e., assigning students to classes according to ability, is a controversial topic among parents, teachers, school administrators, and educational researchers. To determine the effects of inter-class ability grouping, a meta-analytic methodology was applied to 85 studies carried out in elementary and secondary schools. Screening of studies yielded 40 studies at the elementary school level and 45 studies at the secondary school level which were suitable for analysis. Two separate analyses were performed. The first analysis covered achievement results, which were reported in 78 studies. The second analysis covered self-esteem results, which were reported in 24 studies. The results revealed that, while the average achievement effect size was only 0.15, certain types of studies produced stronger results than did other types of studies. The feature that proved most important in determining strength of results was the type of program examined in the study: honors programs designed especially for talented students had clear positive effects when compared to XYZ and remedial programs. Results on self-esteem revealed that the average effect of grouping on self-esteem scores of students was near zero. These findings suggest that homogeneous grouping is often beneficial for talented students, may improve achievement and self-esteem of slow learners, and has little effect on the achievement and self-esteem of average students. Graphs are included. (NRB)