Eighteen studies on the validity of tests of English as a foreign language (EFL) for college admission and placement decisions are briefly summarized. No studies were found on the extent to which these tests reflect language skills, curriculum (content validity), or theory or behavior (construct validity). Instead, all of these studies investigated the predictive validity of test scores relative to some external standard, usually grade point average, but occasionally grades in EFL coursework or indices of decision-making accuracy. Author, affiliation, sample, predictor, criteria, and findings are reported for each study. The Test of English as a Foreign Language, the Graduate Record Examination, the Michigan Test of English Language Proficiency, and other tests were used as predictors, either alone or in various combinations. No predictor explained more than half of the variance in the criterion (grades) and most explained less than one-quarter. This lack of validity may be caused by variability in grading standards, exclusion of test candidates who were not admitted to college, or low content validity of the EFL tests themselves. Two predictors which need investigation are native language achievement and mathematical aptitude. (CP)