This study attempts to analyze the level of students' proportional reasoning in terms of self-efficacy by classifying students' self-efficacy, describing the levels of students' mathematical proportional reasoning, and describing the differences of their levels in terms of self-efficacy. The subjects of this study were students of class VII with a total of 40 students who had received proportion matter. The research instruments are questionnaires, test questions, interviews, observations, and documentation. The data collection procedure begins with students filling out a self-efficacy questionnaire. After the results of the questionnaire were obtained, two students were selected for each category, then reasoning tests and interviews were conducted. The results showed that from 40 students, three self-efficacy classifications were obtained, namely, nine students with high enough self-efficacy, 28 students with moderate self-efficacy, and three students with moderately low self-efficacy. Based on the results of this data analysis, it was found that students' proportional reasoning was at level 0 and level 3. At level 0, students could not solve proportional problems. They can only solve unknown values by using addition, difference, or unpatterned calculation, i.e., using arbitrary numbers and operations. At level 3, students can do proportional reasoning using unit values or scale factors in solving unknown value problems. Then there are differences in the level of mathematical proportional reasoning between students who have high enough selfefficacy, moderate self-efficacy, and moderately low self-efficacy.