This study was designed to compare the complexity, accuracy, and fluency of EFL written text in CLIL and Non-CLIL classes. The study enrolled two groups of undergraduate students from the State University of Malang, Indonesia: an experimental CLIL class (N = 50 students; 22 males and 28 females) and a non-CLIL as control class (N = 50 students; 19 males and 31 females). Students' essays were evaluated quantitatively using some aspects of linguistic proficiency, such as complexity, accuracy, and fluency. The errors were classified as syntactic, morphological, lexical, lexicogram, spelling, and punctuation mistakes. The findings indicated that both CLIL and non-CLIL methods produced comparable complexity, accuracy, fluency, syntactic, morphological, lexicogram, and spelling scores in two groups of students. Meanwhile, for complexity and lexical values, the CLIL and Non-CLIL methods produced significantly different average scores, with the application of the non-CLIL method being higher. On the punctuation variable, the CLIL method can significantly improve the assessment. Several possible explanations for the increase in complexity, accuracy, and fluency include a preference for the English standard, the course's assessment criteria, and practice effects. The findings of this study also provide additional pedagogical implications.
                        
                        
                        
                        
                            
                                Copyrights © 2024