This study contributes to validity studies by determining the cut-off point of an inventory measuring university students' anxiety levels with Angoff, ROC, and Borderline methods and by examining high/low anxiety levels according to these methods point. The study is regarded as a basic research due to the newly-added data in a multi-scoring inventory which validity and reliability studies were previously performed. This study included 290 participants who studied at the Education Faculty of a state university and were planning to take the Public Personel Selection Examination. Trait Anxiety and Test Anxiety Scales were applied to the students. Findings obtained from this study show that the cut-off scores (48 and 48.5) calculated by Angoff and ROC analyses are very close to each other. The consistency coefficient among the judges was examined to prove the reliability of the cut-off points determined by Angoff and Borderline methods. According to the AUC (area under the curve) value obtained in the study, it is seen that the Trait Anxiety Inventory could correctly classify high anxious and low anxious individuals at the rate of 73% (moderate level). It is shown that the scale is quite likely to determine anxiety levels when it is used for screening rather than diagnosis. As a conclusion, the study indicates that when ROC analysis is used to determine the cut-off score of a multi-scoring psychological measuring instrument, such characteristics as sensitivity, specificity, and positive/ negative predictive values that other methods do not have provide more detailed and objective information.