This research was motivated by the problem of low ability to understand mathematical concepts because many students' conceptual understanding results were below the Minimum Competency Standards, especially in mathematics. In connection with this, this research aims to determine the significant average difference between the pretest and posttest and how much the ability to understand students' concepts in learning mathematics about flat figures for class IV at SDN 3 Terban has increased. The research was conducted at SD 3 Terban with a sample of 30 students. The research method used is a pre-experimental quantitative method with a one-group pretest-posttest design. The sampling technique in this research was saturated sampling. Data collection in this research was carried out through tests, interviews, and observations. Data analysis includes a normality test, a Paired Sample T-test, and an N-Gain test. In this test, the Paired Sample T-test results were obtained with a significance of 0.000, which means Ho was rejected and Ha was accepted. The pretest average is 53.6, and the posttest average is 86.7, which shows an increase. Meanwhile, the N-Gain test result of 0.67 means that the increase in students' ability to understand concepts can be categorized as moderate criteria, and the interpretation of effectiveness gets a percentage of 67%, including the quite effective category. Based on these statistical tests, the results obtained can be concluded that the application of the use of CIBATAR learning media significantly impacts the understanding of concepts before and after treatment, and there is an increase in students' understanding of concepts using CIBATAR learning media in mathematics lessons on flat shapes.