Ever since Clarke (1979) first introduced the notion of a language threshold, the threshold hypothesis has attracted many L2 reading researchers' interest. However, none of them has presented any convincing evidence to support the hypothesis. Therefore, this study empirically tests the hypothesis of the threshold model. In addition, using the notion of the threshold level, this study examines the level of L2 proficiency required to read "difficult" texts in the textbook used in College English for Liberal Arts courses in one university in Seoul. Finally, this study explores the relations among L1 reading ability, L2 reading ability, and L2 proficiency. The results of the study show that the threshold level of L2 proficiency exists in L2 reading, and that the data obtained for this study entirely support the threshold model. The findings of the study also suggest that at least 700 in TEPS, 565 in TOEFL, or 770 in TOEIC is needed to read the "difficult" texts in the College English for Liberal Arts textbook effectively. It was also found from the study that the relative contribution of L2 proficiency to L2 reading ability is greater than that of L1 reading ability. Based on the findings, implications for instructional practice are also discussed.