One contributing factor to students’ low understanding of mathematical concepts is their tendency to focus on procedures rather than concepts. This study aims to compare the effectiveness of the Missouri Mathematics Project (MMP) and Problem-Based Learning (PBL) models in enhancing junior high school students’ conceptual understanding in mathematics. This quasi-experimental study used a non-equivalent pretest-posttest control group design involving two 7th-grade classes from a junior high school in Banjarmasin. Class VII A used the MMP model, while Class VII B used the PBL model. Students' understanding was measured using pretest and posttest assessments on ratio and comparison material, and the data were analyzed using SPSS. The results showed that the MMP group had a higher average gain score (0.8741) than the PBL group (0.4722). However, the Independent Samples Test revealed a significant difference (p < 0.001), indicating that both models significantly improved students’ understanding, with PBL showing greater statistical effectiveness. This suggests that although MMP yielded a higher mean gain, the consistent effectiveness of PBL across students makes it more effective in practice. These findings suggest that both MMP and PBL can be applied to strengthen students’ conceptual understanding. Teachers are encouraged to integrate contextual, student-centered strategies such as those in PBL to better support conceptual development in mathematics classrooms.