This study was aimed to find out how English teachers understand the concept of analysis of test items and apply it in constructing test items in Senior High Schools Harapan Persada and Tunas Bangsa Abdya. The targets were two sets of entrance test documents for 2016 and 2017. This research was conducted from 22 to 28 of July 2017. Descriptive analysis design was used as the research method. Interview and checklist were used as the research instruments. The results showed that in Senior High School Harapan Persada for the year 2016 test has 20 test items: 14 (revised), 6 (utilized), and 0 (changed). For 2017 test it has 30 test items: 3 (revised), 27 (utilized), and 0 (changed). Senior high school Tunas Bangsa for the year 2016 has 10 test items: 0 (revised,) 10 (utilized), and 0 (changed). For the 2017 test, it has 15 test items: 15 (utilized), 0 (revised), and 0 (changed). In assessing students at both schools, the maximum score for multiple-choice section is 100; as well as for interview. The final result was gained by calculating both of the sections, devided by two. Meanwhie, in interview test, personal introduction, school domain, daily activities, hobbies and interests, and purpose/decision were used as the measurement. There were differences and similarities in constructing test items of both schools. Senior high school Harapan Persada for the test of 2016 only implemented vocabulary and communicative skills, without including genres and short functional texts. However, for the test of 2017 it has implemented the integrated unit: transactional, interpersonal, short functional texts and genres. Meanwhile, Senior High School Tunas Bangsa (in 2016 and 2017) has implemented integrated unit in arranging English entrance test. It is suggested that the appropriate test construction procedure before providing tests be highly considered by the test developers, particularly English teachers.