<p>The efficacy of academic English instruction in Iranian medical universities is compromised by the inadequate placement of students into English for General Purposes (EGP) courses, resulting in heterogeneous classes with varying proficiency levels. This study investigated the relationship between norm-referenced tests (NRT) of English proficiency and criterion-referenced tests (CRT) of English achievement, assessed the predictability of CRT scores based on NRT scores, examined the forecasting of one CRT (Final Exam) based on another (Midterm Exam), and compared the performance of students grouped by language proficiency versus academic major. This correlational quantitative study involved 220 paramedical students enrolled in EGP courses. The results indicated significant positive correlations (P &lt; 0.05) between NRT and CRT scores, suggesting that NRT scores can serve as a reliable measure for class placement. Further predictive analyses confirmed the feasibility of forecasting CRT scores using NRT data, with midterm test scores identified as strong predictors of final test performance. Additionally, grouping students based on their proficiency level (Scenario B) exhibited greater language acquisition efficacy than grouping them based on their majors (Scenario A). These findings underscore the importance of implementing NRT-informed placement policies in EGP courses to enhance instructional effectiveness and cater to students’ individual language needs. Adopting proficiency-based grouping can enhance learning outcomes, reduce instructional challenges, and ensure a more equitable language learning environment. This study offers practical implications for curriculum designers, educational policymakers, and instructors seeking to refine English for Specific Purposes (ESP) and English as a Foreign Language (EFL) assessment frameworks, particularly in Asian higher education contexts.</p>