This study attempts to determine whether or not ability grouping is desirable for effective instruction. Parents volunteered their children for assignment to a heterogeneous social studies class. Students were divided into a Volunteer Group, a Non-volunteer Group and a Non-response Group. Students for the heterogeneous classes were randomly selected from previous tracks and were distributed among four experimental classes, in a proportion representing the proportion of each group in the eighth grade population. The balance of the Volunteer Group were distributed into regular track classes. Other groups were similarly placed. Results of the factor analyses indicate that student attitudes are not affected by class type. Analyses of variance indicated that students from lower tracks achieved at a higher level in a heterogeneous grouping than in a homogeneous one. Teacher activities were similar with both groups, with student interaction apparently accounting for achievement differences. Sample student questionnaires and tests, as well as parent letters are included. (Author/CJ)