The purpose of this study was to compare reading and math academic achievement scores of a cohort of students who had experienced mixed-age (two-grade span) and multiage (three-grade span) configurations, in a selected nongraded elementary school located in East Tennessee. Student attitude toward school, gender and socioeconomic status were also analyzed. The causal-comparative quantitative approach, along with convenience sampling, was the foundation for this study. Academic achievement normal curve equivalency (NCE) scores from the TerraNova Comprehensive Test of Basic Skills for the 1997-2000 academic years and survey results from the Attitude Toward School Inventory (Meier, 1973) given in the concluding year (2000) were analyzed using ANOVA, ANCOVA and t-tests to determine which configuration produced better results for students. Statistically significant results (p=. 05) were found indicating that the multiage students performed better in reading achievement during the 1997 and 1998 years (ANOVA). ANCOVA results indicated multiage configuration to be statistically significant in 2000 when controlling for prior reading achievement. ANOVA results proved to be statistically significant in math for the multiage configuration in 1998. NCE mean scores in reading and math were higher, some significantly higher, for all four years 1997, 1998, 1999, and 2000 for those students in the multiage (three-grade span) configuration. No statistically significant differences were found in configurations regarding attitude toward school, however in all subtest areas the multiage (three-grade span) students mean scores were higher than the mixed-age (two-grade span) students scores. Findings include a stronger case for multiage (three-grade span) configuration when planning a nongraded developmentally appropriate elementary program.