The effectiveness of an Education Consolidation Improvement Act Chapter 1 program was studied in a large urban school district. A common problem in evaluating remedial programs is that pretest/posttest achievement gains often indicate the effects of the entire curriculum, rather than those specific to the remedial class. In this district, over 50,000 Chapter 1 students were enrolled in more than 250 elementary schools. A conceptually simple method was needed for studying the effects of the remedial program versus the core curriculum. School regressions, residuals, mean residuals for Chapter 1 students at each school, and a number of mean residuals were computed from a database containing pretest and posttest scores. Residuals from within-school regressions were used to model the effect of the entire curriculum. Each school's mean residual indexed the effectiveness of its own Chapter 1 program. Results indicated one-third of the school's mean residuals for Chapter 1 students were positive. A large correlation matrix of variables which might explain the variance patterns was not conclusive. Only a few variables were modestly correlated with the residuals. In addition, few of the explanatory variables correlated with each other as expected. Measuring teacher characteristics might have been helpful in this project. (GDC)