<![CDATA[Mastering mathematical problem-solving is a crucial 21st-century skill, yet many elementary students continue to struggle with abstract concepts. Conventional teaching methods often fall short of bridging this gap, underscoring an urgent need for innovative, concrete instructional approaches. Therefore, this study seeks to assess the effectiveness of the Problem-Based Learning (PBL) model, combined with the congklak game media, in improving elementary school students' mathematical problem-solving abilities, particularly in the KPK and FPB topics. A quasi-experimental method employing a Nonequivalent Control Group Design was utilized. The study included 52 students in Purwakarta, distributed equally into an experimental class receiving PBL with congklak and a control class applying Discovery Learning. Data were gathered through descriptive tests based on Polya's problem-solving indicators and then analyzed using independent-samples t-tests and N-Gain calculations. The results demonstrate a notable improvement in the experimental group, which attained a mean post-test score of 92.15, compared with the control group's 85.81 (p < 0.001). Furthermore, the N-Gain analysis showed that the experimental class outperformed the control class across all problem-solving indicators, with the most substantial differences seen in planning the solution (0.92 vs. 0.72) and implementing the plan (0.81 vs. 0.66). In conclusion, applying the PBL model alongside concrete congklak media significantly enhances students' mathematical problem-solving skills by fostering an active, tangible learning environment. Consequently, educators are encouraged to integrate traditional game-based manipulatives into their pedagogy. This strategy not only deepens mathematical comprehension but also helps preserve local cultural heritage among young learners. Keywords: problem-solving skill, congklak media, problem-based learning, elementary school.]]>
