<![CDATA[Assessment in Arabic language learning today faces a fundamental problem, namely the mismatch between assessment instruments and the language competencies that should be measured. Evaluation practices still tend to be oriented toward cognitive-memorization aspects and have not fully accommodated language skills in an integrative manner. This condition raises academic concerns regarding the validity and accuracy of assessment results. This study aims to conceptually examine the role of item analysis and instrument validity in improving the quality of Arabic language assessment. The research method employs a qualitative approach with a literature study type through descriptive-critical analysis of various literatures related to Arabic language learning evaluation. The results show that the application of item analysis, which includes difficulty index, discriminating power, and distractor effectiveness, as well as testing the validity and reliability of instruments, contributes significantly to improving the quality of evaluation tools. Instruments that meet these principles are proven capable of measuring language competencies comprehensively, covering receptive and productive skills objectively, systematically, and accurately. The implications of this study affirm the importance of reconstructing the Arabic language assessment system based on a scientific and empirical approach to support the development of adaptive, communicative, and 21st-century needs-oriented learning]]>
