The purpose of the study is to explore the effectiveness of reverse engineering pedagogy (REP) and forward project‐based pedagogy (FPP) in K‐12 robotics education. A two‐stage study was conducted in two secondary schools, involving a total of 169 students. Based on the experience of the pilot study (Study 1), we refined the REP and examined its effectiveness through a quasi‐experimental design in the formal study (Study 2), which included four teaching models: the Deconstruction Recovery Model, Troubleshooting Model, Element Minitrim Model, and Structural Innovation Model. Results indicated that students in the REP group performed significantly better and perceived a higher degree of compatibility and creative self‐efficacy than those in the FPP group, which is consistent with the final robotics product evaluation. However, no significant difference was found in the learning attitude towards the course. The research findings highlight the value of REP in robotics education, especially in cultivating creativity and enhance learning performance, however, it needs reasonable planning and design.