Introduction The aim of this study was evaluating if the presence of a humanoid robot could improve the efficacy of a game-based, nutritional education intervention. Study design This was a controlled, school-based pilot intervention carried out on fourth-grade school children (8–10 years old). A total of 112 children underwent a game-based nutritional educational lesson on the importance of carbohydrates. For one group (n = 58), the lesson was carried out by a nutritional educator, the Master of Taste (MT), whereas for another group, (n = 54) the Master of Taste was supported by a humanoid robot (MT + NAO). A third group of children (n = 33) served as control not receiving any lesson. Methods The intervention efficacy was evaluated by questionnaires administered at the beginning and at the end of each intervention. The nutritional knowledge level was evaluated by the cultural-nutritional awareness factor (AF) score. Results A total of 290 questionnaires were analyzed. Both MT and MT + NAO interventions significantly increased nutritional knowledge. At the end of the study, children in the MT and MT + NAO group showed similar AF scores, and the AF scores of both intervention groups were significantly higher than the AF score of the control group. Conclusions This study showed a significant increase in the nutritional knowledge of children involved in a game-based, single-lesson, educational intervention performed by a figure that has a background in food science. However, the presence of a humanoid robot to support this figure's teaching activity did not result in any significant learning improvement.