Abstrak Penelitian ini bertujuan untuk mengetahui apakah ada perbedaan kemampuan representasi matematis siswa yang diajarkan menggunakan pendekatan pembelajaran Realistic Mathematic Education (RME), Contextual Teaching And Learning (CTL), dan konvensional. Jenis penelitian ini adalah penelitian ekperimen semu. Populasi penelitian ini adalah seluruh kelas VIII SMPN 24 Bengkulu Utara dengan sampel yaitu siswa kelas VIII A sebagai kelas eksperimen I, siswa kelas VIII B sebagai kelas eksperimen II, dan siswa kelas VIII C sebagai kelas kontrol. Data dikumpulkan dengan tes kemampuan representasi matematis siswa  kemudian  di analisis dengan uji Anava. Hasil analisis uji Anava menunjukkan adanya perbedaan yang signifikan antara kemampuan representasi matematis siswa yang diajar melalui pendekatan RME dengan pendekatan CTL, maupun dengan pendekatan konvensional. Hasil penelitian menunjukan bahwa kelas yang diajar menggunakan pendekatan pembelajaran RME diperoleh rata-rata skor akhir kemampuan representasi matematis siswa sebesar 16,8. Kelas yang diajarkan dengan pendekatan pembelajaran CTL diperoleh rata-rata skor akhir kemampuan representasi matematis siswa sebesar 17,9. Sedangkan dikelas yang diajarkan dengan pembelajaran konvensional diperoleh rata-rata skor akhir kemampuan representasi matematis siswa sebesar 13,80. Hasil uji BNT (Beda Nyata Terkecil) diperoleh bahwa perlakuan yang memberikan hasil yang berbeda adalah perlakuan antara eksperimen I dengan kontrol dan eksperimen II dengan kontrol.   Abstract This study aims to determine whether there are differences in the mathematical representation abilities of students who are taught using the Realistic Mathematic Education (RME), Contextual Teaching And Learning (CTL), and conventional learning approaches. This type of research is quasi experimental research. The population of this research was all class VIII SMPN 24 Bengkulu Utara with a sample of students of class VIII A as the experimental class I, students of class VIII B as the experimental class II, and students of class VIII C as the control class. The data were collected by means of a mathematical representation ability test of the students and then analyzed using the Anava test. The results of the Anava test analysis showed that there was a significant difference between the mathematical representation abilities of students taught through the RME approach with the CTL approach, as well as the conventional approach. The results showed that the class that was taught using the RME learning approach obtained an average final score of students' mathematical representation ability of 16.8. Classes taught with the CTL learning approach obtained an average final score of students' mathematical representation abilities of 17.9. Meanwhile, in the class taught by conventional learning, the final average score of students' mathematical representation ability was 13.80. The results of the LSD test (Least Significant Difference) showed that the treatment that gave different results was the treatment between experiment I and control and experiment II with control.