While children can and do learn STEM (science, technology, engineering, and math) skills from interactive media, like smartphone and tablet applications, it is not clear why digital interactivity might benefit children’s learning. As new interactive technologies, like virtual reality, become more popular, it is important to understand the mechanisms behind learning from these media to best design them for learning. Some research demonstrates that interactive media may not produce greater learning than non-interactive media (such as watching a video), and these findings come in contrast to social theories of learning. Another theory that is untested in this area of research is the Limited Capacity for Motivated Media Message Processing which would suggest that working memory demands could hinder children’s learning from interactive media. The current study was an experiment (N = 188, children ages 7-12) in which children completed both puzzles in virtual reality and physical puzzles to demonstrate learning spatial skills. Working memory demands were manipulated through working memory precues that were predictive (children saw a picture of a completed puzzle for 3s) or neutral (children saw a picture of a blank square or rectangle for 3s). Findings demonstrate that the working memory precue manipulation did not relate to number of physical puzzles completed after using virtual reality, but there was a main effect of age such that older children completed more physical puzzles than younger children. Theoretical and practical implications are discussed.