The purpose of this study was to compare five methods of computing school effectiveness indices (SEIs) from longitudinal data. The methods were within-school regression, within-school regression corrected for the unreliability of measurement, mean difference scores, average individual residual scores, and school residual scores. The sample consisted of 3,769 third-graders from 70 elementary schools in the Midwest. The raw data consisted of Total Reading scores from the Metropolitan Primary II Achievement Test administered in fall 1970 and spring 1971. While the various school effectiveness indices differed from one another and in their correlations with other variables, little evidence could be found for the lack of validity of any school effectiveness index. Further, all of the school effectiveness indices were highly stable across samples, except for the indices for initially high-scoring students. Finally, predictions from nonlongitudinal data furnished reasonable estimates of school effectiveness as measured by one of the indices. (Author/CK)