This study aims to determine the effect of the Realistic Mathematics Approach (RMA) on students' mathematical problem-solving abilities and to compare the learning completeness of students taught with RMA and conventional learning. The type of research used is a quasi-experiment with a post-test only control group design. The research sample consisted of students in class VIII-1 as the experimental class who were taught using the RMA, and students in class VIII-2 as the control class who were taught using conventional methods. The research instruments were pre-test and post-test questions in the form of essays compiled based on indicators of mathematical problem-solving abilities. Data collection techniques were carried out through tests. Before testing the hypothesis, the data was tested for normality using Shapiro-Wilk and homogeneity using Levene's test. The test results showed that the data of both groups was normally distributed and homogeneous. Data analysis used the Independent Sample T-test. The t-test results showed that the t-count value was greater than the t-table (6.822 > 2.024), so the null hypothesis (H₀) was rejected and the alternative hypothesis (Hₐ) was accepted. Thus, it can be concluded that there is a significant effect of the use of the Realistic Mathematics Approach on students' mathematical problem-solving abilities. Students who learn with RMA have better problem-solving abilities than students who are taught through conventional learning at SMP Daya Cipta Medan. In addition, the mastery of problem-solving skills among students in the experimental class was higher than that of the control class, indicating that the realistic mathematics approach is more effective in improving the quality of mathematics learning.