This study examines the changes in teaching practices during the implementation of a pedagogical model called the mathematics reasoning approach (MRA), which was founded on 2 critical areas in mathematics, problem solving, and writing to learn. Three algebra teachers implemented the approach with their classes, which were divided into control (traditional) and treatment (student-centered approach) groups, with their levels of implementation measured by the Reformed Teaching Observation Protocol (RTOP) instrument. The implementation of the model resulted in positive changes in teachers' pedagogical practices and yet showed the struggles they faced. In the control classes (traditional classes), the level of teaching remained the same throughout the study, whereas in the treatment classes (MRA classes), a significant improvement in the level of teaching was observed. Even though the teachers' implementation levels differed from each other, the patterns of the change for all 3 teachers showed similarities; that is, their MRA teaching level was initially equivalent to their control teaching, but as the semester progress, there was a separation between the 2 approaches. This separation between the 2 approaches appeared to be dependent upon the pedagogical area on which the teacher put emphasis. The skill of questioning seemed to be 1 which each teacher first changed.