It has been argued that, compared with the traditional graded form of school organization, the nongraded form is superior in developing pupil classroom behavior, attitudes, and achievement that are related to generally accepted educational objectives. To test the validity of this view, multivariate analyses of covariance were performed on one nongraded experimental group of 224 pupils and two traditional graded control groups totaling 483 pupils, all from the K-6 age range and divided into normal age, underage, and overage groups for purposes of analysis. Results indicate that (1) the nongraded form of organization encourages development of conceptual maturity and participation in group activities; (2) teachers in nongraded schools tend to be more accepting of disorderly pupil behavior; (3) the graded organization seems to encourage pupil development in achievement, attitude toward school, and contributing activities during teaching episodes; (4) overage pupils in the nongraded school seem to be more contributing members of their classes than overage pupils in graded schools; (5) underage pupils generally scored highest and overage lowest on the measures used; and (6) the research design seems appropriate for use in the evaluation of experimental programs. (TT)