Many scientific concepts and theorems are often abstract and challenging to relate to real-life problems, making it difficult for students to grasp them. Therefore, some researchers have attempted to enhance students' understanding by employing a contextual learning approach, which allows students to apply scientific knowledge to real situations in their daily lives. The aim is to improve students' learning experiences by moving away from rote memorization. However, if a contextual gaming approach is offered without encouraging deep reflection, students may focus solely on the game itself and overlook the importance of fully understanding the knowledge and contemplating the meaningful relationships between scientific concepts. To address this issue, for this study we developed a Concept Mapping-based Digital Game-Based Learning for Complex Chemistry Problems (short for CM-DGBL-CCP) learning system to assist students in understanding complex chemistry problems. To verify the effects of the proposed approach, the experiment was conducted in a secondary school with two groups. The experimental group with 49 students adopted the CM-DGBL-CCP learning model, while the control group with 56 students utilized the traditional digital game-based learning for complex chemistry problems (T-DGBL-CCP) learning model. The experimental results revealed that there were no significant differences between the two groups of students in terms of learning achievement and cognitive load. However, the experimental group students outperformed the control group in areas such as problem-solving tendency, scientific self-efficacy, scientific learning strategies, and the ability to use deep-level strategies to solve problems.