In 1986, a study was conducted by the Rancho Santiago Community College District (RSCCD) to evaluate the relationship between reading placement scores and classroom performance. The study sample consisted of students from fall 1985, spring 1986, and fall 1986 who had been tested using the College Board Assessment and Placement Test upon registering. An analysis of the relationship between course grades, the percentage of students who were successful, and placement scores indicated that there was only a modest relationship between placement scores and grades and that a fair number of students who had reading scores below the 25th percentile were successful (grade of "C" or better) in transfer-level courses. These findings suggested that placement scores were not a reliable predictor of classroom success for many courses and challenged the predictive-validity of such measures. A second part of the study utilized a series of discriminant analyses to determine the ability of reading scores to differentiate between successful and unsuccessful students in transfer-level courses. The results showed that only 46% to 79% of the students could be correctly classified on the basis of reading placement scores alone. Based on study findings, it was concluded that the use of cut-off scores was inappropriate since many students who would have been predicted to fail actually earned a grade of "C" or better. An appendix contains a series of tables of individual courses sampled with Spearman correlations and grade distributions. (Author/UCM)