A non-experimental design was used to determine if scores of students enrolled in specified major coaching schools were significantly higher than scores of comparable uncoached groups. Score increases at two Scholastic Aptitude Test (SAT) coaching schools and Law School Admission Test (LSAT) schools were compared. Over 1,400 SAT examinees and 31,000 LSAT examinees, administered these tests between 1974 and 1977, were involved. Comparisons were reported by study subgroup: uncoached first-time examinees; coached first-time examinees; uncoached second-time examinees; second-time examinees coached before the first test; and second-time examinees coached between test administrations. Results indicated that coaching was significantly effective in raising SAT scores. LSAT coaching was marginally effective, possibly because of the abnormally large control group increases and the relatively low correlation between grade point average and LSAT scores. However, the Federal Trade Commission acknowledges that there are "several major flaws in the data analysis, making the results unreliable." (The history of both tests, an analysis of coaching research; a description of the sales, tuition costs, and claims of major coaching schools; extensive graphs of test score data; and case histories of 17 coached LSAT examinees are included). (CP)