Data from the College Board Validity Study Service show that the average multiple correlation of the Scholastic Aptitude Test (SAT) with college grades peaked in 1974 and then tended to decline. Data from other sources also estimate a small average decline from 1974 to 1985. This study documented changes in the SAT and related these changes to trends in the predictive validity of the SAT, focusing on changes in test format, content specifications, statistical characteristics of the test, and equating procedures associated with SAT test forms taken by classes graduating from high school in 1971 through 1985. Data came from November and December test forms for these years. Analyses indicated that the slight decline in SAT predictive validity was not due simply to the shortening of the SAT, nor to changes in content or statistical characteristics, nor to changes in equating methods. No patterns of change in the various indices that were reviewed were consistent with the patterns in validity. Thirty-nine tables in the text, 17 in an appendix, and 14 figures illustrate these analyses. Appendixes present SAT directions and sample questions and contain supplemental tables. (Contains 33 references.) (SLD)