Penelitian ini bertujuan untuk mendeskripsikan efektivitas model pembelajaran Relating, Experiencing, Applying, Cooperating, Transferring (REACT) berbantuan Game Edukasi dan Reciprocal Teaching berbantuan Game Edukasi ditinjau dari kemampuan pemecahan masalah matematis siswa. Penelitian ini adalah quasi experiment dengan posttest only control design. Populasi dalam penelitian ini adalah siswa kelas VIII MTs Sultan Fatah Gaji Demak tahun ajaran 2017/2018 dan kelas VIII-A, VIII-B, dan VIII-C dipilih sebagai sampel penelitian. Hasil penelitian menunjukkan bahwa (1) terdapat perbedaan signifikan rata-rata kemampuan pemecahan masalah matematis antara siswa yang mendapatkan pembelajaran dengan model REACT berbantuan Game Edukasi, model Reciprocal Teaching berbantuan Game Edukasi, dan model pembelajaran konvensional (ekspositori); (2) kemampuan pemecahan masalah matematis siswa yang belajar menggunakan model pembelajaran REACT berbantuan Game Edukasi lebih baik dari pada siswa yang belajar dengan pembelajaran konvensional; (3) kemampuan pemecahan masalah matematis siswa yang belajar menggunakan model Reciprocal Teaching berbantuan Game Edukasi lebih baik dari pada siswa yang belajar dengan pembelajaran konvensional; (4) pembelajaran dengan model REACT dan Reciprocal Teaching berbantuan Game Edukasi sama-sama efektif ditinjau dari kemampuan pemecahan masalah dengan rata-rata kemampuannya signifikan di atas 70; dan (5) terdapat pengaruh positif keaktifan siswa pada kelas yang belajar menggunakan model REACT dan Reciprocal Teaching berbantuan Game Education terhadap kemampuan pemecahan masalah matematis siswa. The effectiveness of the Relating, Experiencing, Applying, Cooperating, Transferring (REACT) and reciprocal teaching model assisted by educational gameAbstractThis research aimed to describe the effectiveness of Relating, Experiencing, Applying, Cooperating, Transferring (REACT) learning models assisted by Educational Game and Reciprocal Teaching assisted by Educational Game in terms of students’ mathematical problem-solving abilities. This study was a quasi-experiment with a posttest only control design. The population in this research was students of 8th-grade of MTs Sultan Fatah Gaji Demak, Indonesia in the academic year 2017/2018 and class VIII-A, VIII-B, and VIII-C were selected as research samples. The results showed that (1) there was a significant difference in the average mathematical problem-solving ability between students who received learning with the REACT model assisted by Educational Game, the Reciprocal Teaching model assisted by Educational Games, and conventional learning models (expository); (2) the mathematical problem-solving ability of students who learn using REACT learning models assisted by Educational Games was better than conventional learning; (3) the mathematical problem-solving ability of students who learn using the Reciprocal Teaching model assisted by Educational Game was better than conventional learning; (4)The learning with the REACT and Reciprocal Teaching models assisted by Educational Games was equally effective in terms of problem solving skills with an average ability significantly above 70; and (5) there was a positive correlation between the activeness of students in class learning using REACT and Reciprocal Teaching models assisted by Game Education on students’ mathematical problem-solving abilities.