Eiken tests are extremely popular within Japan, and are considered to be English proficiency tests like TOEIC and TOEFL. However, these tests have been criticized in regard to how reliable they are in providing a true measure of overall English proficiency and in regard to the quality of the test items they utilize for measuring that proficiency. This paper outlines an analysis of the content and quality of test items within a Third Level Eiken test, given to 119 first year students at a national technical college in a regional area of northern Japan. Data gathered from student responses to these test items was used to calculate a reliability estimate of this test for comparison with that consistently gained in TOEIC tests using the same formula. Item discrimination and item facility values were also calculated to determine the degree of difficulty of test items and the degree to which each test item separates good test performers from the not-so-good performers. Analysis of the data gathered from this study group revealed a low reliability estimate for test items in comparison to that normally obtained for the TOEIC test, and that 79% of test items should be revised or deleted because they discriminate poorly between test takers, or are too easy or difficult for the target test takers.