Analysis of question items is very important in learning evaluation because it determines the quality of the evaluation tool used. The aim of this research is to analyze the results of the final semester Arabic evaluation at the Al-Ma’arif Junior High School. The researchers conducted a series of tests consisting of measurements of validity, reliability, differential strength, difficulty level, and debilitating power. The method used in this research is quantitative descriptive method. The population of the study was 25 students of the 7th grade in the IT High School Building Amal School Year 2022/2023. As for the sample used, the final evaluation of the Arabic semester of the teaching year 2022/2023 was documentation and observation. The result of this study is; in the validity test found 30% of questions can be said to be valid. Then the remaining 70% is declared invalid because it has a significance value <0,05 and the value of R count is smaller than R Table. Then in the reliability test found using Cronbach's Alpha method, and the element of this question is worth 0.625 with the number of questions (N of items) amounting to 40. Difference shows 6 questions in the category "Bad once in distinction, 25 questions in category "badly once in differentiation", 6 issues in the categories "Sufficient enough in distinguishment", 3 questions with category "Good once in distinctness" whereas with the question of category "good once in discrimination" was not found. Difficulty level there are 25 questions with very easy questions category, 14 questions with easy categories, 1 questions with difficult categories. The fractional strength of the answer option A indicates that the question is not working well with the student as the data indicates more fractionals of the question are rejected than accepted. The same applies to options B, C, and D. Thus, a question that does not meet the revised criteria can be concluded or a new question created and tested again. This research can be used as evaluation material for teachers to create quality evaluation tools. Further research can focus on discussing the development of superior evaluation instruments.