This study aims to see the effectiveness of the use of mathematical pyramid media in learning media courses to improve the academic achievement of fourth semester students of the Mathematics Education Study Program, FKIP UNA, Academic Year 2019/2020. The sample of this research is two classes, namely class IVA (experimental class) and class IVB (control class). The pretest mean value of the experimental class (64.82) with the highest value of 82 and the lowest score of 46, the pretest value of the control class (61.48) with the highest value of 78 with the lowest score of 42. From the pretest homogeneity test there was no difference in variance or the two samples homogeneous where Fcount <Ftable (1.86). From the pretest data normality test, it was obtained that L0 for the experimental class (0.1274) and L0 for the control class (0.117). For post-test data, it is obtained L0 for the experimental class (0.1587) and L0 for the control class (0.1082). The value of L0 is smaller than L table (0.161) so it can be concluded that the data is normally distributed. The average post-test score of the experimental class (87.9) with the highest score of 90 and the lowest score of 63, while for the control class the average post-test score (80.6) with the highest score of 90 and the lowest score of 54. = 2.11> ttabel = 2.01 which means that Ha is accepted.
Keywords: Mathematics pyramid media; academic achievement; learning media