Despite the rapid development of motion-based technologies, their role in embodied learning remains ambivalent. This study aimed at investigating the potential added value of motion-based technologies in the context of geometry elementary education. An explanatory sequential design was adopted, composed of two phases. First, an experiment was conducted: students in the experimental group (n = 15) participated in a digital embodied intervention while students in the comparison group (n = 16) participated in a non-digital embodied intervention, to learn about angles. Quantitative results showed that, while students' cognitive load was not differentiated, students in the digital intervention outperformed their counterparts in terms of learning gains and emotional engagement. Qualitative analysis of interview data allowed better understanding of students' engagement and cognitive load supporting the quantitative findings. The authors reflect on these findings and the implications for the use of motion-based technologies in embodied interventions for learning in mathematics.