This study examines the relationship between knowledge of academic vocabulary relevant to an American university setting and academic success as measured by grade point average (GPA) of undergraduates who were non-native speakers of English at an university in the United States. To study this relationship, this study addresses issues in developing a list of vocabulary items relevant to undergraduate study in the United States; developing, administering and scoring an appropriate test of vocabulary size; and interpreting the results of the test. The study also addresses the reliability of a checkbox test, the usefulness of common word families is estimating vocabulary size, the usefulness of nonwords as indicators of overestimation of knowledge by test takers, and the rate at which nonwords may be included in a checkbox test of vocabulary.;The results do not show a clear relationship between knowledge of academic vocabulary and academic success as measured by GPA; however, the results do tend to indicate that the lower limit of a successful student's vocabulary size may be as low or even lower than 2700 word families. The results also indicate that the checkbox test can be reliable, that common word families are not very useful in estimating vocabulary size for non-native speakers of English enrolled as undergraduates, that nonwords are significant indicators of overestimation of knowledge, and that nonwords may be included at a low rate (e.g.10%) on a checkbox test of vocabulary provided there are a sufficient number of nonwords.