This study aims to determine the difference in the improvement of students' mathematical problem-solving ability using the Problem Based Learning model with the conventional model in terms of Adversity Quotient types of quitters, campers, and climbers, and to determine the interaction between the Problem Based Learning model and Adversity Quotient on improving problem solving ability. This study uses quantitative research methods with a sample of 72 grade X students at SMA Negeri 73 Jakarta. This sampling technique uses cluster random sampling technique. The instruments used were Adversity Quotient questionnaire and problem-solving ability test. Data analysis techniques include validity test, reliability test, normality test, homogeneity test, N-gain test, sample t-test, and two-way anova test and post-hoc test. The results of this study show that (1) there is no difference in the improvement of mathematical problem solving ability between students who learn Problem Based Learning model and conventional model, (2) there is a difference in the improvement of students' problem solving ability in terms of Adversity Quotient types of quitters, campers, and climbers, (3) there is an interaction between Problem Based Learning model and Adversity Quotient on the improvement of problem solving ability. The post-hoc test shows that there is an interaction between climbers and quitters and quitters and climbers