THE USE OF AMERICAN COLLEGE TEST (ACT) SCORES AS A CRITERION FOR ADMISSION TO INSTITUTIONS OF HIGHER LEARNING PROVIDED PREDICTIVE INFORMATION, BUT ALSO HAD LIMITATIONS. BEGINNING IN 1961 THIS STUDY FOLLOWED 1,062 FRESHMEN OF THE UNIVERSITY OF MISSISSIPPI THROUGH THEIR 4 YEARS OF COLLEGE. THE ACT WAS GIVEN TO ALL INCOMING FRESHMEN SEPARATED INTO THREE GROUPS BASED ON THEIR ACT SCORES. GROUP I CONSISTED OF 102 STUDENTS WITH COMPOSITE STANDARD SCORES OF LESS THAN 15. GROUP II CONSISTED OF 250 STUDENTS SCORING BETWEEN 15 AND 18. GROUP III CONSISTED OF 709 STUDENTS SCORING HIGHER THAN 18. CORRELATING COMPOSITE SCORES OF GROUPS I AND II WITH CUMULATIVE GRADE POINTS AT THE END OF EACH SCHOOL YEAR RESULTED IN CORRELATIONS VERY CLOSE TO ZERO. SIGNIFICANT CORRELATIONS, HOWEVER, WERE OBTAINED FOR GROUP III. WHEN SCORES WERE CONSIDERED FOR ALL INCOMING FRESHMEN, THE CORRELATION BETWEEN COMPOSITE ACT SCORE AND ACADEMIC ACHIEVEMENT WAS RELATIVELY STRONG AND POSITIVE. TEST LIMITATIONS WERE EVIDENT SINCE DROPOUT RATE AND ACT SCORES WERE NOT HIGHLY CORRELATED. THE HIGHEST DROPOUT RATE WAS AMONG STUDENTS SCORING RELATIVELY WELL ON THE ACT. FURTHERMORE, THE TEST WAS UNACCEPTABLE IN EVALUATING THE ABILITY OF FOREIGN STUDENTS TO SUCCEED IN COLLEGE. OVERALL, THE ACT TEST PROVED TO BE A GOOD INDICATOR OF ACADEMIC SUCCESS OF INCOMING FRESHMEN. (GB)