The aim of this research is to comprehensively describe the influence of a realistic mathematics approach based on local wisdom on elementary school students' mathematical numeration abilities. The research method used was quasi-experimental, with a posttest control group research design design. The population in this study were all state elementary school students in class V odd semester of the 2022/2023 academic year in Ternate City. The sample involved 124 students spread across two elementary schools based on experimental and control levels and classes. The instruments used include mathematical numeracy ability test items and student attitude scale guidelines. Data analysis was carried out quantitatively and the statistical tests used were the t-test, Mann-Whitney test , and two-way ANOVA test. Next, to find out the average increase in students' mathematical numeration abilities, namely by using the mastery level and normalized gain test (N-Gain). Based on the results of data analysis, it shows that learning factors have an influence on increasing students' mathematical numeration abilities, namely where the normalized gain test (N-Gain) was obtained at 0.58 with a class average value of (62.82%), but there was no interaction between the learning approach and the students' KAM (initial mathematical ability) category. When viewed from Meltzer's classification, the overall increase in students' mathematical numeracy abilities is included in the high category. The conclusion is that there is an increase in the achievement of each indicator of students' mathematical numeracy abilities through a realistic mathematics approach (RML) based on local wisdom (LW) which is better than conventional/usual learning.