Changes in the predictive validity of the Scholastic Aptitude Test (SAT) that have been observed in the University System of Georgia (USGA) over the past two decades were studied. Several analyses were performed using student data from 27 USGA schools to determine the effectiveness of the SAT in predicting freshman grades. National trends indicated that the validity coefficients for high school record (HSR) and HSR+SAT declined from 1964 to 1982. Findings from Georgia also demonstrate an appreciable decrease in validity coefficients for the past 6 or 7 years. The decrease in SAT validity coefficients is accompanied by similar variations for high school average (HSA) coefficients and for multiple correlation coefficients involving combinations of HSA and SAT. The same general trends were observed when annual variations were smoothed by computing 5- and 10-year averages. These general trends were affected by the variability of student abilities, as measured by the SAT, and academic performance, as reflected in high school and college grades. It is cautioned that decreases in the magnitude of validity coefficients do not mean a decline in the meaning and significance of the SAT as a measure of verbal and mathematical abilities. Appendix A contains 5-year averages in validity coefficients, and Appendix B contains profiles of cooperating institutions. (Contains 9 figures, 15 tables, 27 figures in Appendix A, 23 figures in Appendix B, and 7 references.) (SLD)