This research is a Classroom Action Research (PTK) consisting of 2 cycles, each consisting of 2 meetings. Before giving action, an initial ability test is first given and at the end of each cycle is given a problem-solving ability test. From the results of the analysis of initial ability test data, it was obtained that many students achieved learning completeness from the initial ability test, namely 18 out of 28 students or 21% with a grade average of 26,7. The results of data analysis in the first cycle after applying the Reciprocal Teaching type cooperative learning model showed that the number of students who achieved learning completeness was 13 out of 28 students or 61% with a grade average of 70,7. The results of data analysis at the end of cycle II with the same learning obtained the number of students achieving learning completeness, namely 24 out of 28 students or 86% with a grade average of 90,86. This means that there is an increase in students' mathematical problem-solving skills from cycle I to cycle II. Based on classical completeness criteria, this learning has reached the classical completeness target.