This research aims to find out whether problem-based learning is effectively used to increase students' mathematics learning self-efficacy. This research was carried out in the even semester of the 2023/2024 academic year with a total of 24 students. The method used in this research is descriptive analysis. The instrument in this research used a questionnaire. The students' mathematics learning self-efficacy questionnaire consisted of 35 questions consisting of positive questions and negative questions, using a Likert scale. In trial I the average result of the students' mathematics learning self-efficacy questionnaire was at a percentage of 74.96%, then in trial II the results of the students' mathematics learning self-efficacy questionnaire increased with an overall average score of 86.47%. Based on the results of trial I and trial II, students' self-efficacy in learning mathematics using problem-based learning increased by 11.51%. Thus, it can be concluded that problem-based learning is effectively used to increase students' mathematics learning self-efficacy.