This Study investigated the validity and reliability of Modified Teacher-Made Biology Junior Corticated of Education (JCE) Mock Test Items which had diagrams, pictures and tables tailored for learners with visual impairment in selected secondary schools within the South West Education Division of Malawi. The primary purpose of the study was to determine if these modified test items, maintained the validity and reliability of the original items. Findings indicated significant discrepancies: 70% of the modified items were less reliable and valid than the original versions. The inconsistency was evident in the difficulty levels of the modified items, with some items proving more difficult and others less difficult than the original items. Additionally, item characteristic curve analysis revealed that some modified items were overly challenging even for learners without visual impairments. Reliability analysis highlighted poor performance in some items, with low discrimination indices (e.g., QM10 = 0.08, QM19 = 0.19). Content validity assessments demonstrated that some modified items did not measure the same concepts as their original items for example; they were derived from different topics and objectives. For instance, question QM20 was not aligned with the original item's topic and objective. The study concluded that most modified test items failed to maintain the original difficulty range, making them unreliable. The findings underscored the need for comprehensive guidelines and enhanced in-service training to develop valid and reliable test items for learners with visual impairments, ensuring equitable assessment practices.