This study examines the alignment between mathematics learning outcomes and assessment practices in three Moroccan universities. The purpose is to evaluate whether assessment tests adequately represent the intended learning outcomes and reflect the results of teaching activities. The study focuses on the Algebra 2 module, which serves as a reference framework for the targeted learning outcomes. The research adopts a psychodidactic approach, combining the analysis of assessment questions and the module’s specification table. The representativeness of the intended learning outcomes in each test is measured using the chi-square goodness-of-fit technique. The results reveal significant inconsistencies in the representativeness of the content and cognitive levels. The findings indicate a recurrent overrepresentation of algebraic reasoning, as well as variations between years and institutions. The high chi-square values (ranging from 45.65 to 125.32) suggest that several tests do not adequately cover the intended learning outcomes. These findings point to a misalignment between assessment practices and the learning outcomes framework. This study highlights the need to rethink university assessment practices by adopting a systematic approach based on clearly defined learning outcomes and psychometric principles. Such improvements could contribute to more valid, balanced, and equitable evaluation of students’ competencies.

