<![CDATA[The low ability of students to transform mathematical problems into verbal, visual, or symbolic representations approached that teacher centered learning and the lack of linking the material to real contexts. This research investigates how Realistic Mathematics Education (RME) integrated with STEM activities influences students' ability to represent mathematical concepts. This research adopted a quantitative approach with a quasi-experimental pretest-posttest control group design to evaluate the effectiveness of the implemented learning approach. The sampling of participants was carried out through cluster random sampling. The instrument employed for data collection comprised interview guides, observation sheets, validation forms, and test items designed to measure students' mathematical representation ability. Data collection was carried out through observation, interviews, and mathematical representation ability tests. Data were analyzed using normality tests, homogeneity tests, independent sample t-tests, paired sample t-tests, N-gain scores, and Cohen d effect size. The results follow that RME learning based on STEM activities in the experimental group was better with a high N-gain score (0.73) compared to the control class with a medium N-gain score (0.54). The Cohen's d effect size analysis yielded a value of 1.43, falling into the "very large" category. These results confirm that RME instruction combined with STEM activities successfully enhances students' mathematical representation ability.]]>
