The promotion of interest in Science, Technology, Engineering and Mathematics (STEM) areas is a priority in the current and future context of Mexico, evidenced through concrete actions, both public and private. As such, there is a need for tools to accurately measure vocational interest to effectively promote STEM education. According to Holland Occupational Themes theory, vocational interest can be conceived as an aspect of personality. There is empirical evidence supporting personality prediction using neurophysiological signals. Therefore, vocational interest may be predicted using similar data. The objective of this study is to develop an intelligent system that estimates engineering interest in children through their physiological response to engineering-related activities. The participants were 13 children between 6 and 15 years old. For each participant, 8 different electroencephalographic channels were measured, as well as the electrodermal activity, heart rate variability, facial gesticulation, and body temperature data for four 2-hour sessions of engineering-related extracurricular educational activities. Psychometric tests were also included to evaluate the children’s interest in specific engineering fields according to their learning activities. The generated data was processed and used to design a machine learning algorithm. The results and implications are discussed.