In this paper, we explore the impact of problem statement readability and lexical richness on cognitive load during programming exercises. Cognitive load theory suggests that the complexity of instructions can significantly affect student performance, especially in tasks requiring high mental effort. Using NASA-TLX, this study measures students’ cognitive load working on JavaScript and React programming problems with varying readability levels. Results show that students with higher programming proficiency tend to experience lower cognitive load, even with more complex problem descriptions, while less experienced students benefit from simplified instructions. This paper provides recommendations for instructional design to optimize cognitive load based on student proficiency, that could be incorporated in a learning environment.