This study was designed primarily to evaluate the effectiveness of three multi-age classes at Carnarvon School, Vancouver, B.C. (each class having an age range of three years: ages 6-8, 7-9, 9-11) in developing children's self-concepts and attitudes toward school. No significant difference was found between the multi-age and regular classes in mean raw scores on the Self-Concept Scale and Pupil Behaviour Inventory employed in the study. In addition, the study aimed to evaluate the effectiveness of the innovation in promoting individualization of instruction and in encouraging children to help each other with school work. The study also sought the opinions of teachers and children in the multi-age classes regarding the innovation. The opinions of the teachers of the two younger multi-age classes were very positive; but the teacher of the oldest multi-age class felt that the innovation was ineffective. All three teachers felt that special care should be taken in the placement of children in multi-age classes. The children's opinions were consistent with the teachers' opinions in most matters pertaining to the innovation. Finally, the study attempted to determine whether multi-age classes achieved at least as well as regular classes in basic reading and arithmetic skills. Reading achievement scores from the Gates-MacGinitie Test were obtained at the Grade 5 level, and these data were analyzed with I.Q. scores from the Henmon-Nelson Test as a covariate. Arithmetic achievement scores were obtained from the Vancouver Surveys at the Grade 3 and Grade 6 levels. The results are provided. (Author/DB)