<![CDATA[Item analysis is a tool to find out whether the question is sufficient with a good question category or not. In addition, the purpose of item analysis is also to help improve the test through revision or removing ineffective questions, as well as to find out diagnostic information to students whether or not they have understood the material that has been taught. In this case, there are still a lot of student scores that are still below the KKM standard. There is no analysis of the problem so that one will not know whether the question is said to be a good question or not. Meanwhile, the teacher only corrects right and wrong. then not tested first for various reasons. That's why the researcher wanted to analyze the madrasa exam questions made by the teacher himself. In this research, the researcher uses quantitative descriptive method, which is a research that describes the data analysis in numbers. The researcher found that the relationship between questions and language skills included 2 language skills, namely: the ability to "read" 52.5% and "write" 47.5%, included in the sufficient category with a percentage of 50%. The questions that are in accordance with the Basic Competence (KD) include 3 KD, namely KD "3.1" 57.5%. KD “3.2” 20%, and KD “3.3” 22.5%, are in the good category with a percentage of 75%. The quality of the exam questions from the aspect of validity and reliability. Which includes valid questions are 28 questions. Meanwhile, 12 questions that are not valid. The questions made by the teacher have high reliability with the number 0.711 > r table 0.254. The test items are viewed from a higher level of reasoning. The test items showed the level of cognitive domain, C1 "remembered" got 37.5%, C2 "understood" 20%, C3 "applied" 22.5%, C4 "analyzed" 20%. The questions that show the HOTS level are 20%, while the questions that show the LOTS level are 80%.]]>
