This paper summarizes research with the American College Testing Program (ACT) at one university over a 3‐year period. Subjects were 4,597 freshmen divided into 5 groups by college and sex. General positive conclusions were that the ACT (a) is a good measure of general scholastic aptitude, (b) has definite administrative advantages to the university by providing scores prior to a student's admission, and (c) aids prediction for certain subgroups of students. However (a) the tests are probably not difficult enough and have a restricted range of scores, (b) the battery is inefficient in terms of student time, and (c) the subtests do not have differential validit