An attempt was made to relate college or university student performance with high school effectiveness. School effectiveness was measured in terms of the college performance of graduates, after controlling for academic aptitude. The data base consisted of freshmen at selected institutions of a large state university system (N=150,000). Steps in the analysis included: (1) developing multiple regression equations for each institution by gender and minority status; (2) predicting college grade point average (GPA) through these equations, using as independent variables the proportions of free and reduced price lunches, the Scholastic Aptitude Test (SAT) verbal and mathematics scores, and college credit hours carried and earned; (3) using the appropriate regression equation for the institution to calculate the residual (the difference bewteen actual and predicted GPA); and (4) determining the average residual for graduates of each high school to rank them on the basis of mean residuals. Results showed a modest pattern of suburban and urban high schools tending toward lower college performance than expected; on the other hand, there were high schools located in rural areas that had positive mean residual GPAs. This suggests that some rural and small-town schools do produce students who perform at a higher level than would be predicted on the basis of their SAT scores or the socio-economic status of their high schools. This method may have promise for extending assessment possibilities; methodological implications are discussed. One flowchart is provided, and an appendix lists the multiple regressions by institution. (SLD)